Chapter 12

1917-18. Smith-Barry’s instructional methods begin to influence attitudes towards all aircrew training.

While the RNAS had been refining its overall flying training sequence, the RFC had been introducing a far more methodical approach to the specific business of teaching people to fly. Concerned at the lack of ability displayed by many of the replacement pilots he was being sent while commanding No 60 Sqn in the autumn of 1916, Maj R Smith-Barry had analysed their failings. He used his conclusions to develop a training philosophy predicated on the use of an aeroplane with appropriate handling qualities and fitted with dual control, the Avro 504, and a comprehensive syllabus to be taught by experienced pilots who had themselves first been taught how to teach (see Note 10 to Chapter 7). While the need for such an approach may seem self-evident today, it was a revolutionary development at the time.

The so-called ‘Gosport system’, was initially applied on a limited scale, producing an immediate and remarkable improvement in both the competence and the confidence of newly qualified pilots. The success of Smith-Barry’s methods led to a major restructuring of the RFC’s training organisation, a project which had been under active consideration since February 1917. In July these deliberations were concluded by the granting of approval for the establishment of seven new flying training schools on a trial basis. These Training Depot Stations (TDS) proved to be a great success and between March and August 1918 the remainder of the pilot training system was converted wholesale.

The necessary resources for forming new TDSs were obtained by disbanding the existing Training Squadrons and abandoning the long-established expedient of employing Service Squadrons as makeshift training units prior to their mobilisation, many of the latter also being disbanded.¹ Thereafter most TDSs provided ‘all-through’ flying training, the initial phase on Avro 504s and the advanced phase on a current operational type, thus combining the functions of latter day basic and advanced flying training schools with the addition of some aspects of an operational conversion course.²

The introduction of TDSs was accompanied by corresponding improvements in the instruction of rear crews. Four factors drove the changes to this element of the system. First, as a spin-off from Smith-Barry’s philosophy, there was an increased awareness of the need to improve the quality of the instruction being provided for all aircrew, which led to a reappraisal of some elements of the observer training sequence. Secondly, there was the need to cater for the range of skills required which broadened at an accelerating rate during 1918, leading to the establishment of a series of new schools to deal with new specialisations and tactics. Thirdly, as the nature of the air war changed, due to the introduction of more capable aeroplanes and more sophisticated equipment, it was necessary constantly to revise the length and content of courses in response to feedback from the front line squadrons. Finally, the merging of the RFC and RNAS meant that the use of the separate, and in some instances duplicated, schools inherited by the new Service on 1 April had to be rationalised to produce an integrated RAF system.

It was envisaged that once the reorganisation of the training system was complete new operational squadrons would be formed two months before they were scheduled to be deployed overseas. The pilots for a headquarters element and three flights were to be contributed by four different TDSs and transferred to a designated Mobilisation Station. There they would be joined by any necessary rear crew members and a draft of ground tradesmen, issued with their operational equipment and commence their work-up programme.

The changes outlined above were very extensive and they could not be implemented overnight. Moreover, the process was prolonged by the fact that a change to the syllabus of one school could have a knock-on effect on that of another. Some idea of the instability within the system may be gauged from the rather bewildering series of redesignations to which some units were subjected, not to mention the cumbersome titles which they acquired.³ The reorganisation of the training system may be considered to have begun as early as the summer of 1917 but, despite gathering pace markedly following the establishment of the RAF, it was still not really complete when the Armistice brought the process to an abrupt stop. Nevertheless, by that time the system had more or less achieved its overall objective in that the deployment of the first of a new generation of freshly formed operational squadrons manned by comprehensively trained personnel was about to begin.4 At least, the crews should have been comprehensively trained, but some doubt must remain as to the extent to which this aim was actually being achieved (see pages 123-25).

1916-18. The contribution of the RFC’s overseas training organisations.

The changes which affected the training system were not confined to UK-based units. As early as April 1916 it had been decided to set up a training organisation in Egypt, primarily to produce additional pilots for the war in Europe, and this began to function in July. The schools in Egypt were originally provided with drafts of trainees from home but the expansion of the UK-based system during 1917 rendered it less necessary to sustain the time-consuming practice of shipping people to and from the Middle East. Thereafter the training units in Egypt began to draw the bulk of their intake from local internal recruiting and from South Africa, much of their output being absorbed by the growing number of in-theatre squadrons. Policy changed again in 1918, however, and before the war ended the training facilities in Egypt, which had been considerably expanded and reorganised to resemble the late-war UK system, was once again being provided with large batches of cadets from home.

The first training unit in Egypt to have any potential relevance to observers was No 3 SoMA which opened at Aboukir in November 1916, this being supplemented in December by the nucleus of a School of Aerial Gunnery. By the time that the latter was formally established in the following April it was already turning out some fifteen officers per week. Most of these were pilots and EOs specialising in armaments, however, and very little real effort was devoted to the formal instruction of observers until 1918.

Operational units arriving from the UK (Nos 14, 17 and 47 Sqns) had generally brought a full complement of observers with them, these reflecting a variety of levels of skill and experience. Most of the observers who succeeded those in the original drafts, and those required for locally raised units, were trained in-house at squadron level, using similar methods to those employed in France in 1915-16. Very few trained replacements were sent out from home, indeed, from late 1917 onwards, the net flow was in the opposite direction, with observers trained in the Middle East being posted home for assignment to squadrons in France.

Serious consideration was not given to the needs of back-seaters in the Middle East until as late as November 1917 when an Artillery Observation School was set up at Almaza. This unit inducted its first course of pilots in January 1918 and its first batch of observers in February. By the time of the Armistice it had turned out 200 observers and 321 pilots, plus 115 W/T operators and six EOs qualified as wireless officers.

Paralleling developments at home (see below) an Armament School had been established at Abassia and in May 1918 a flight of four DH 6s of the Artillery Observation School was hived off to form the nucleus of No 3 School of Navigation and Bomb Dropping (SNBD). It was not until August that this school came into its own, however, moving to Helwan on the 19th where it was expanded to have a fleet of twenty-three aircraft, BE2es, BE12s, AW FK3s (and/or FK8s?) and the DH 6s. With Maj G Merton in command, training began in September and by the end of the following month No 3 SNBD had graduated a total of twenty-four students and had a further forty-six pilots and thirteen observers in residence. In September, again mirroring the UK-based training organisation, the School of Aerial Gunnery was merged with the Aerial Fighting School at Heliopolis to form No 5 Fighting School. Some idea of the size of this unit’s throughput can be gauged from its turnover during October in which month alone it took in 155 trainees and graduated forty-nine pilots and forty-eight observers.

Despite the considerable expansion and increasing sophistication of the training system in Egypt, the regional requirement for observers remained relatively small and, so far as two-seater units were concerned, the RFC’s overseas Order of Battle had more or less stabilised by early 1918. Furthermore, since there were few combat losses the rate of turnover was also low, compared to that in France. Once a unit was up to strength therefore, it needed relatively few replacements. Some idea of the numbers of back-seaters required in the ‘side-show’ theatres can be gained from the routine strength return for the week ending 14 March 1918.⁵ This reported on the availability of a total of 119 observers and seventy-one balloon officers.6 Two of the former were with the Artillery Observation School; the distribution of the remainder is shown at Figure 12. No 26 Sqn had been withdrawn at the beginning of February but it is worth noting that, until then, despite the apparent lack of any formal establishment for them, there had been several additional observers serving in East Africa.⁷ To give some scale to these figures, the 14 March return had also reported that there were 447 qualified pilots serving in these theatres with another 1,031 in transit to Egypt, already undergoing training there or awaiting disposal after having gained their ‘wings’.

Fig 12. Distribution of balloon officers and observers in overseas theatres (excluding France and Italy) as at 14 March 1918.

Anticipating in some respects the Empire Air Training Scheme of WW II, another extensive network of schools had been set up in Canada early in 1917.⁸ The system in Canada differed from the one in Egypt in that it was not provided with trainees from the UK, all intakes being locally recruited in North America. The training organisation in Canada was chiefly intended to produce pilots but it did eventually turn its hand to observers; the first batch graduated in July 1918 and before training ceased 137 had qualified, 85 of whom had been sent overseas.⁹ By that time the units dealing with back-seaters were a Cadet Wing at Long Branch, No 4 School of Aeronautics (SoA) in Toronto, an Armament School at Hamilton, the School of Aerial Gunnery and the School of Aerial Fighting at Beamsville and the School of Artillery Co-operation at Leaside.¹⁰

An RE8 of the Palestine-based No 142 Sqn in 1918. (Chaz Bowyer)

1918. Role-specific establishments are introduced and the twenty-four-aircraft corps reconnaissance squadron concept begins to be realised.

By the autumn of 1917 the War Office had concluded that the current one-size-fits-all squadron establishment had outlived its usefulness and that it was time to replace it with a series of documents tailored to meet the individual requirements of squadrons operating in six increasingly diverse roles.¹¹ In the meantime, to reflect the fact that the planned expansion of corps squadrons to a strength of twenty-four aircraft was in abeyance (see page 71), the current universal eighteen-aircraft establishment had been amended to delete the six NCO gunners leaving just the twenty commissioned observers.¹² When the first editions of the role-specific establishments were published at the end of January 1918 that for a corps reconnaissance squadron, now restored to a notional strength of twenty-four aircraft, still allowed for only twenty officer observers.13

Beginning in 1916, the RFC’s aeroplanes were identified by a unit marking – a combination of lines/bars or a geometric shape, eg a triangle, square, disc, hexagon. On 22 March 1918 (the day after the Germans had launched Operation MICHAEL, their final offensive on the Western Front) all units, other than fighter squadrons, were ordered to remove their markings. Thereafter most corps reconnaissance squadrons applied a simple ‘number-in-unit’. Since some progress began to be made towards achieving the long-approved target strength of twenty-four aircraft per squadron, identities up to 24, and occasionally even higher, began to appear later in the year. This is an RE8 of No 16 Sqn. (P S Swain)

Perhaps because the RFC had been temporarily operating on an eighteen-aircraft-per-squadron basis during the winter, no one appears to have noticed the error at the time but the real requirement was obviously going to be for twenty-six observers – eight per flight plus two for the headquarters. Someone finally picked this up in March, by which time the winter moratorium had been lifted and the long postponed expansion of squadrons to their intended strength of twenty-four aircraft had begun to get under way. GHQ duly requested that a further six officer observers be added to the establishment of a corps reconnaissance squadron.¹⁴ The increase in strength was approved but the extra six men were to be NCOs, not officers.¹⁵

Based on the number of pilots on strength, by the time of the Armistice only five corps squadrons were operating at the twenty-four aircraft establishment; four were at twenty-one, the other eleven were still at eighteen, but none had their full complement of observers (see Fig 21 on page 128).¹⁶

1918. The impact of the capabilities of second-generation aeroplanes on the functions of the Corps Observer.

While the home-based training machine had been undergoing a major overhaul and expansion there had been significant feedback from the two-seater squadrons operating in France. By the spring of 1918 the RFC had accumulated a year’s operating experience with its RE8s and FK8s and Ludlow-Hewitt’s earlier contention, that pilots were best suited to handle fire control, had gained widespread acceptance throughout the corps reconnaissance community. Although many units had already put this philosophy into practice, however, it had not yet been endorsed by officialdom as being the preferred operating procedure. Nevertheless, all pilots destined for corps squadrons had been passing through the Artillery and Infantry Co-operation School at Winchester since the autumn of 1917 and, since they attended the full course alongside observers, it was clear that there was considerable duplication of effort.

Furthermore, it had been decided as early as the previous July that the Bristol Fighter would be introduced for corps work from April 1918. The anticipated imminent arrival of these much more capable aeroplanes prompted the operators to review the way in which they conducted their business. There was a growing consensus (at middle management level at least) that if the responsibilities of crew members were not clearly redefined and assigned, much of the Bristol’s considerable potential would be wasted on artillery work.

Lt-Col I A E Edwards, OC 15th Wg, submitted his thoughts on the matter to HQ V Bde in March 1918.¹⁷ He went so far as to claim that recent experience had shown that it was actually ‘a danger for the observer to carry out a shoot’, maintaining that it was essential for him to concentrate on look out and rear defence. It followed that observers therefore needed to be expert gunners and, while acknowledging that they ‘had shot down a few (Huns)’, Edwards considered that many opportunities for combat victories were being missed. Although he did not actually make the (surely even more important) point, an extension of his argument would suggest that many aeroplanes were also being unnecessarily lost. In short, Edwards, like Ludlow-Hewitt, believed that pilots were actually better than observers at gunnery direction, but he accepted that they could do it properly only if they were totally confident that their rear defence was in safe hands.

The nub of the problem was that corps observers were currently being trained to carry out those aspects of the job which pilots now wanted to do, indeed were doing, at the expense of developing their gunnery skills which were sadly lacking – or so the pilot fraternity maintained. The latest thinking was that the occupants of rear cockpits needed to be trained to look up, rather than down, and to be able to shoot straight when necessary. On the other hand, while they still needed to have some understanding of the business of fire control and be able to handle a camera, they no longer required an exhaustive knowledge of artillery procedures nor did they need to be quite so handy with ‘the buzzer’. Edwards’ letter had included specific suggestions as to how he thought the training of corps observers should be tailored in the future.

Having gained the immediate approval of V Bde, Edwards’ ideas were promptly forwarded to HQ RFC and thence to the Air Ministry with the endorsement of the GOC, by now Maj-Gen J M Salmond. A revised training sequence was introduced in April. Thenceforth, following their indoctrination course at a School of Aeronautics, corps observers were to undergo a fortnight of academic gunnery instruction at the Armament School at Uxbridge,18 followed by three weeks of practical air-to-air and air-to-ground gunnery at Hythe and/or New Romney before spending just two weeks at Winchester. The planned intake into this system was to be at an initial rate of twenty-five per week.

The anticipated introduction of the Bristol Fighter in large numbers led to a reappraisal of the employment of observers in the spring of 1918. The relaxed attitude of the crew of this one, which belonged to No 62 Sqn, is a little surprising as the engine is running – and the cockpit appears to be empty. (J M Bruce/G S Leslie collection)

In the event, problems with the supply of engines seriously disrupted production of the Bristol Fighter. Apart from a handful operated by autonomous Long Range Artillery Flights, one of which was allocated to each brigade towards the end of the war, only a few Bristols were made available for corps reconnaissance work, leaving the ‘Harry Tates’ and ‘Big Acks’ to fly doggedly on until the Armistice.¹⁹ In the increasingly hostile air combat environment of 1918, however, the corps observer’s enhanced ability to fight off attackers became an invaluable asset. It is true that a handful of artillery pilots did manage to score some victories, but they were exceptions to the rule. The more usual situation was neatly summarised by No 6 Sqn’s Sgt Eddington (see Chapter 6, Note 24):

‘The pilot (of an RE8) had a Vickers gun which synchronised through the propeller. But nine times out of ten it didn’t – it hit the propeller. They were wooden propellers and the result was tremendous vibration because of the imbalance. You never used your gun – you relied on the observer who had a Lewis.’

1918. The increasing importance of the rear gun.

While it was clearly necessary for corps reconnaissance back-seaters to be expert gunners, it was equally important that they should be so within the growing force of day bomber squadrons where many rear cockpits were occupied by non-commissioned personnel. This was in sharp contrast to corps reconnaissance units, of course, where there were not supposed to be any NCOs at all at the turn of 1918 (see Note 12)

Apart from tactical work in direct support of a ground offensive, daylight raids by DH 4s and DH 9s were usually carried out in formation. This tactic served to focus the effects of the attack but it also denied any freedom of manoeuvre and thus rendered the fixed forward-firing Vickers guns of the pilots virtually useless. On the other hand, flying in formation concentrated the defensive crossfire which could be delivered from the rear cockpits, the survival of bombers being, therefore, almost entirely in the hands of observers and gunners.

The evolving situation in the fighter reconnaissance world was more complex.²⁰ Because the 1½ Strutter of 1916 had provided its pilots with the innovation of a forward-firing gun, there had been a natural tendency for them to want to use it. One of the first to appreciate the implications of this was Louis Strange, the man appointed to oversee gunnery training at Hythe at the end of 1916, who wrote:²¹

‘I went back to England fully convinced that the gun mounting, ie the machine, was far more important than the gun. In spite of this, I realised that, given the speed, a good two-seater fighter is bound to be a better machine than a single-seater Scout, and anyway, the RFC could not do without observers, who had to be trained in gunnery and a host of other things. Every man who goes into the air on a fighting machine is a gun-layer first and last, and so his success, and the success of his Squadron and the whole of the RFC depended on gun-laying ability.’

He was right, of course, but as previously discussed (see page 22), while the two-seater Sopwith may have been the right ‘machine’ in 1916 it had soon been outclassed and by the spring of 1917 it had been accepted that its best chance of survival lay in flying in formation. Nevertheless, if a formation was broken up it inevitably became every man for himself. This was only a realistic proposition, of course, if the two-man crew could learn to fight as a single unit. This was difficult to achieve in a Sopwith, as graphically described by No 45 Sqn’s Frank Courtney:22

‘Our two guns, one forward and one aft, were intended to be a menacing combination, but they turned out to be not even a combination; for, while the pilot was weaving the plane around in order to use his fixed gun, the poor gunner in the rear had little chance to take effective aim from an unpredictably swerving platform largely obscured by a swinging and bobbing tail. So only one gun or other could be effective at the same time, and what we had was two independent gunners, back to back, neither of whom knew what the other was trying to do, and each of whom wondered what was going on when he heard a burst of fire from the other end of the plane.’

A DH 4 of No 27 Sqn (above) and a Bristol Fighter of No 1 Sqn AFC (below). Note that in the DH 4 the crew stations are some six feet apart, making communication, let alone co-operation, very difficult. In the Bristol the pilot and observer/gunner sat back-to-back, which made for much greater efficiency as a fighting team. (Both J M Bruce/G S Leslie collection)

What was clearly needed was a decision as to which gun should be afforded priority when a two-seat fighter was obliged to fight alone and, to this end, in April 1917 the CFS had carried out a series of trials involving a Bristol Scout and a Sopwith 1½ Strutter (sometimes flown solo), each attacking and being attacked in turn. Analysis of the camera-gun film demonstrated quite conclusively that in all cases it was a mistake for the pilot of the two-seater to concentrate on trying to bring his own gun to bear, since ‘infinitely superior results’ were obtained if he manoeuvred to give his observer the best possible firing opportunity.²³

Similar work had been going on at the RFC’s Experimental Station at Orfordness, using, in addition to the Bristol Scout and the Sopwith, examples of the more capable RE8, Bristol Fighter, DH 4 and AW FK8, engaging with a Sopwith Triplane and a BE12. While not expressed in quite such positive terms, the report on the Orfordness project had also acknowledged the need for a two-seater pilot to fly with a view to providing his observer with a clear shot.²⁴ The trial had also convincingly demonstrated (if this were still necessary) that, regardless of the type of aircraft and/or its role, mutual co-operation between the crew members was the key to success in any two-seater. In view of this, it is a little disconcerting to find that, according to 2/Lt V Voss, replacement Bristol Fighter pilots were still being sent to France as late as February 1918 without ever having flown with a passenger, let alone a trained observer.²⁵

To operate the Bristol Fighter successfully in air combat obviously demanded a team effort between co-equal partners, regardless of their individual ranks and status. Some idea of the scale of the contribution made by back-seaters can be gained from the record of No 20 Sqn. The official early post-war accounting credited this squadron with a total of 613 successful claims, although much later research raised this to 619. Leaving aside the nuances of whether the enemy aircraft had been destroyed, captured, driven down out of control or merely damaged, an analysis of the records of these engagements suggested that only 138 of the victories claimed had been attributable to pilots, the remaining 481 (78%) having fallen to the guns of observers or gunners.²⁶

Folk knowledge tends to associate success in the field of air combat with squadrons which flew single-seat Camels and SE5as. It may come a surprise to some, therefore, to learn that that the most successful, ie the top-scoring, British fighter unit of the war was the exclusively two-seater equipped No 20 Sqn and that it was not the only two-seater squadron to establish a remarkable combat record.

Despite the level of damage which they inflicted on the enemy, however, the activities of the relatively pedestrian two-seaters have never excited the same level of public interest as have the exploits of the Camel drivers and others of their kind. Furthermore, such attention as they have attracted has, by association with their more glamorous colleagues, tended to focus on their pilots. Sadly, those who flew and fought with them are often treated as little more than footnotes. Thus, while the names of a handful of two-seater pilots, like McKeever, Latimer, Thompson and Atkey, still have some limited currency, those of notably successful back-seaters, like Powell, Noel, Fletcher and Gass, are virtually unknown.

While the annals of the RFC have little to say about observer officers, the deeds of its non-commissioned gunners were afforded even less attention. Arch Whitehouse summed up the overall situation as follows:²⁷

‘You will not find any non-commissioned man listed among Great Britain’s aces of World War I. Furthermore, few commissioned officers were considered important enough to be included in the glorious company of aerial aces. The two-seater fighter squadrons had many such heroes, but they sported the winged-O of the Poor Bloody Observers and were not fit company for the full-winged heroes.’

As a back-seater himself, Whitehouse was clearly biased (and it must also be acknowledged that his writings do tend to hyperbole), but in this instance it is difficult to take issue with him. It was not until as late as March 1918 that the RFC bothered to revise its method of recognising victories to acknowledge formally the part being played by back-seaters. The new rules made it clear that, ‘When an enemy machine is brought down by a two-seater both the pilot and the observer are credited with 1 EA.’²⁸

Fig 13. The distribution of non-pilot aircrew serving with RFC squadrons in France in mid-January 1918. At this stage all observers were officers. Aerial gunners could hold ranks up to sergeant-major but the vast majority were corporals or below.

1918. Changes in the employment of non-commissioned back-seaters.

Oddly enough, in view of the crucial importance of aerial gunnery, at the close of 1917 the War Office decided to dispense with aerial gunners. This paradox was more apparent than real, however, because the manpower allocated to a squadron by this time envisaged the provision of an observer for every aeroplane so gunners had actually become surplus to requirements. Furthermore, the category of the aerial gunner had been introduced to distinguish these men from observers at a time when all of the latter were additionally qualified in airborne photography and trained to work with the artillery and to use wireless equipment. This was no longer the case, of course, because army observers had not been required to pass the Brooklands/Winchester course for several months, their professional training now being undertaken entirely at Hythe/New Romney. Since, apart from their status, there was now no significant difference between a commissioned army observer and a non-commissioned aerial gunner, there was no longer any need to maintain the previous distinction.

It remained to decide which of the two labels should be applied. While it might have been more appropriate to have regarded all of these men as gunners, ‘the Brass’ would no doubt have wished to avoid the presentational difficulties that were bound to have arisen had they tried to ‘downgrade’ officers to work at what was by now widely perceived to be a non-commissioned trade. This approach would also have caused significant administrative complications by introducing yet another RFC grade – that of the commissioned aerial gunner. The line of least resistance was to opt for the alternative solution and ‘upgrade’ gunners to become NCO army observers.

While these arguments are fairly self-evident, they do not appear actually to have been articulated at the time, at least, not publicly. The specific catalyst which had provoked the demise of the aerial gunner had been provided by Trenchard, who had been invited to comment on the allocations of manpower and transport contained within the drafts of the projected role-specific unit establishments (see page 90). HQ RFC’s response included the succinct recommendation, ‘All Aerial Gunners to be abolished.’²⁹

London concurred and the establishment of a Service Squadron (still the current, one-size-fits-all variety) was duly amended in late December 1917. A follow-up letter explained that the six aerial gunners currently on the notional strength of each army two-seater squadron were to be replaced by NCO observers ‘who will be trained in the same way as Officer Observers’. The six aerial gunners on corps squadrons were to be withdrawn and not replaced.³⁰

Although unit establishments had previously provided for up to six ‘serjeant’ gunners per squadron (see page 38), until now most of them had actually flown as air mechanics or corporals. To underline their enhanced status, all noncommissioned back-seaters were now going to wear, at least, three stripes, the amendment noted immediately above having used the term ‘NCO Observer (Sergeant)’.

As with all amendments to personnel policy, it took time for these changes to percolate down to the staffs and the term ‘aerial gunner’ took several months to fade away. It still featured, for instance, in unit establishments published on 29 January 1918, for squadrons stationed in France, and on 16 February for those in the Middle East.³¹ Because their qualifications and/or experience did not permit all of them to be automatically remustered as sergeant observers, it is possible that a few gunners may have lingered on the books until the end of the war (see Chapter 15, Note 10).

Figure 13 provides a summary of the actual manning levels within the force of two-seater squadrons based in France in January 1918.32 Already substantially down on the peak strength reached in the previous summer (see page 40), the deletion of aerial gunners from all establishments meant that their numbers declined sharply thereafter. In addition to those tabulated, there were a further three qualified RFC observers flying Handley Pages with A Sqn, RNAS at Ochey and five probationary gunners on the strength of the Aerial Gunnery Range at Cormont. Oddly enough, although they were equipped with single-seaters, Nos 43 and 60 Sqns were each also carrying a solitary gunner on their books at this time.³³

As an aside, Figure 13 makes an interesting comparison with Figure 10 (on page 40). The manning levels in both cases were on the generous side, but remained in the same ballpark – a grand total of 780 back-seaters (262 qualified) for a force of thirty-four two-seater squadrons in March 1917 compared to 852 (367 qualified) for only thirty-two squadrons in January 1918.³⁴ Of more significance, however, was the completion of the re-equipment programme that had just been getting under way when the first of these sets of statistics was being compiled. In the course of just nine months the old BE2s, Moranes, Nieuports and Sopwiths had all disappeared to be replaced by more modern types. Of the previous generation, only the FE2 soldiered on, its activities now largely restricted to the hours of darkness.

Another return, similar to that represented by Figure 13, but raised in early April 1918 (by which time there were four more multi-seat squadrons in the line³⁵), reflected the revised manning policy. Since aerial gunners were no longer a factor, the report confined itself to the distribution of observers of which there were the following totals:³⁶

a. Qualified officers	346
b. Probationary officers	319
c. NCOs (qualifications not specified)	49
Total	714

At the time, and in accordance with the recent changes to establishments, all of the NCOs were confined to units flying Bristol Fighters or DH 4s. This constraint was not to remain in force for long, however, because, as part of the run-up to the formation of the RAF, the Air Ministry had conducted yet another review of the (by now role-specific) establishments of all RFC squadrons. The revised editions were published with an effective date of 1 April 1918.³⁷ The extracts at Figure 14 reflect the latest thinking on the ratio of officer to sergeant observers at this stage of the war. Note that it was intended to make quite extensive use of NCOs in heavy bombers and that they had recently been re-admitted to corps squadrons (as recorded on page 91). Neither of these innovations ever really caught on, however, and by mid-July the manning of most Handley Page squadrons, ie those operating with the Independent Force, had been further refined to provide a total of twelve officer observers and only four sergeants.³⁸ The employment of sergeants on corps reconnaissance duties was never exercised to any great extent either; of the 137 NCO observers on the strength of the fifty-four two-seater units operating under the aegis of HQ RAF at the end of the war only nine were flying with corps squadrons (see Figure 21 on page 128).³⁹

Fig 14. The officially approved establishments of officer and NCO observers for RAF squadrons in France as at 1 April 1918, an asterisk indicating those roles in which it was considered acceptable to employ additional NCOs if insufficient officers were available.

Although there was ample provision for NCO observers to fly in Handley Page 0/400s, by the summer of 1918 most were actually officers. (Chaz Bowyer)

As a footnote to the demise of the aerial gunner, it should be recorded that the trend towards specialisation became increasingly marked over the next few months and by October it had been decided to reintroduce dedicated NCO gunners, although this intention was to be frustrated by the Armistice.

1918. The RFC reconsiders its policy towards NCO pilots

While this book is intended to focus on non-pilot aircrew, in order to maintain some sort of perspective, it has been necessary to pay some attention to pilots and, in particular, to the way in which they were trained. Since space has already been devoted to non-commissioned gunners and observers, in order to maintain the balance, it is appropriate to consider the RFC’s attitude towards the employment of sergeant pilots.

Provision for non-commissioned pilots had been made from the very outset. As early as 12 April 1912, the White Paper which had published the arrangements for the formation of the RFC had noted that each of the planned seven squadrons was to have had twelve aeroplanes and twenty-six pilots; only half of these pilots were to have been officers. The first non-commissioned pilot, Cpl Frank Ridd (26), gained his RAeC Certificate on 4 June 1912. Others were to follow but their numbers never actually kept pace with those of officers so that by the time that war was declared officers who had qualified as pilots outnumbered those without commissions by almost five to one.⁴⁰

Furthermore, in the light of experience, the RFC had refined its requirements by mid-1913 so that it now employed what were known as First and Second Class pilots, the qualification standards being published in September.⁴¹ In practice, the majority of non-commissioned personnel progressed no further than becoming Second Class pilots, which meant, in essence, that they had passed the tests associated with the RAeC Certificate and had demonstrated an awareness of only some of the technical aspects of aviation. First Class pilots had to have passed examinations in a much broader range of topics at the CFS, in addition to having accumulated an ‘adequate’ number of flying hours.

Even so there were marked distinctions within the First Class classification, in that officers had to pass in all eight subjects examined at Upavon, whereas non-commissioned pilots were not tested on the theory of flight or on meteorology. Furthermore they were not required to demonstrate a knowledge of troop formations or an ability to identify warships, nor, beyond an ability to map-read and use a compass, were they required to be familiar with aerial reconnaissance procedures and techniques. Whatever its initial intentions may have been, therefore, it is quite clear that by 1913 the RFC can already have had no serious intention of using its non-commissioned pilots operationally and, although there were a few exceptions, this policy remained essentially unchanged throughout the war. Until 1917 there was no prohibition on the training of noncommissioned pilots but the majority of those who did manage to qualify were assigned to second-line units where they served as, for instance, ferry pilots at Aircraft Parks or staff pilots at the School of Aerial Gunnery.

A return of all officers and aircrew serving with the BEF in September 1915 noted only two NCOs, both of them with No 3 Sqn.⁴² A similar headcount made in late April 1916, but reflecting all officers and aircrew serving with the RFC, shows thirty-one non-commissioned personnel being carried on the strength of UK-based units as pilots, twenty of them still undergoing training (and most of the eleven who were qualified were graded as only Second Class), but there were still only three NCO pilots flying with operational units in France – one each with Nos 1, 3 and 5 Sqns.⁴³ Despite the considerable expansion of the corps over the next two years, although more NCOs were trained, they continued to represent only a tiny proportion of the total number of pilots available.⁴⁴

In mid-1916 it was ruled that all holders of Second Class Certificates would have to qualify to First Class standard within six months or forfeit their certification. Several concessions were subsequently granted; for example, the original retrospective clause within the new regulation was removed so that pilots already holding Second Class Certificates needed to re-qualify only at that level, the obligation to upgrade being confined to newly qualified pilots. Similarly, the six-month limit was waived for any Second Class pilots already serving with an Expeditionary Force.45

Three NCO pilots, Sgts L G S Gadpaille, H N Lee and C K Robinson, but they were never very numerous in the RFC and few of those who did gain their ‘wings’flew on operations. (J M Bruce/G S Leslie collection)

Nevertheless, the thin end of a wedge could be discerned here and the evident reservations over the utility of the two-tier system would eventually lead to questions being asked about the real value of having any non-commissioned pilots. In this general context the Bailhache Committee made three significant statements in November 1916: ‘In the early days of the war some 50 non-commissioned officers became pilots…..’ and ‘Every pilot must now be an officer. There are a few exceptions.’⁴⁶ Unfortunately, the last two statements plainly contradict each other. What the writer was presumably trying to convey is that, despite the RFC’s overwhelming preference for officers, a handful of NCO pilots was still being accepted, in addition to the relatively small numbers that were already on strength.

The employment of sergeant pilots on operations peaked during 1917. There were, for instance, a total of twenty-seven of them on the strength of the squadrons serving with the BEF in March,⁴⁷ twenty-five in May⁴⁸ and twenty-eight in August.⁴⁹ All of these NCOs were overborne against each unit’s current establishment which provided at that time for a total of twenty-one pilots per squadron, all of whom were to be commissioned – one Squadron Commander and two ‘spare’ Flying Officers with the headquarters, and a Flight Commander plus five Flying Officers for each of three flights. By August 1917 there were forty-seven squadrons and one independent flight in France, reflecting an overall requirement for 992 pilots. There were actually 995 on strength, but only 967 of them were officers, the remaining twenty-eight being sergeants, which represented a ratio of officers to NCOs of the order of 35:1.⁵⁰

Since there was no official establishment for sergeant pilots, it would seem likely that the primary function of the handful serving in France may have been to provide a cushion to ensure that a squadron would remain fully operational if there were any delay in providing replacement officers. This is not to say that NCO pilots were used only as makeweights. Those squadrons which had sergeants on strength certainly made full of use of them, although it is worth noting that very few NCO pilots were carried on the books of single-seat fighter squadrons until (as discussed below) the summer of 1918. Of thirty-two NCOs at the front in the late summer of 1917, for instance, one was flying Pups and three Nieuports; all of the others were on two-seaters.⁵¹

Long before this, however, the RFC’s practice of employing (almost) exclusively officer pilots had become policy. In June 1917 the War Office announced that, with immediate effect, ‘the training of non-commissioned officers and air mechanics as pilots will be discontinued’ (thus removing any lingering doubts as to the accuracy of the previously noted observation contained in the Bailhache Report). ‘Special cases’, should there be any, would still be entertained but, in general, all successful applicants for flying training were now to be posted initially to a Cadet Wing for eventual commissioning on the General List. This regulation was not to be applied retrospectively, but any serving NCO pilot could apply to become an officer; if recommended, he was to be directly commissioned in the field without having to go through the cadet stage.⁵²

Sergeants never represented much more than 3% of the available pilots in France and, despite the continued expansion of the RFC, the universal commissioning policy meant that their numbers actually began to decline from the autumn of 1917. By January 1918, there were only sixteen of them, this figure contrasting markedly with that for noncommissioned back-seaters of whom there were no fewer than 230 (see Figure 13 on page 95)⁵³

For a time, it looked as if things were about to change again because, within a few weeks of the RFC’s announcing that it did not intend to employ any more NCO pilots, it had been obliged to reconsider this decision. There were two reasons. First, aviation was not the only field in which high grade manpower was required. The terrible casualty rate in the trenches meant that competent and capable young men with leadership qualities were urgently needed by other branches, particularly the infantry, and it was being argued, in some of the corridors of the War Office, that the RFC’s 100% officer pilot policy could be sustained only at the expense of the rest of the Army.

The second reason was the July 1917 decision to double the size of the RFC. While it might have been realistic to expect to be able to find enough soldiers with the potential to permit the ‘old’ RFC to operate on an all-officer basis, it was very doubtful whether this ideal could be realised with the newly projected 200-squadron RFC. The staffs foresaw a recruiting crisis and, although it would involve a reversal of policy, reinstating NCO pilots, in comparatively large numbers, would provide an obvious solution to the problem.

Before implementing such a sweeping change in policy it would clearly be necessary to ascertain whether it would be acceptable to front-line commanders and in August the War Office submitted an outline proposal to the CinC BEF and sought his opinion on it.54 Initial calculations indicated that, considering only the forty-four squadrons currently serving with the BEF, it might be possible to replace as many as 290 of the commissioned pilots with sergeants. Such a substitution would also save 143 batmen although this might have to be partially offset by the provision of a number of cooks and waiters to cater for the increased numbers of NCOs.⁵⁵

Having first consulted with Trenchard, Field Marshal Haig eventually agreed to the introduction of significant numbers of NCO pilots with the proviso that the measure should be tried ‘purely as an experiment, subject to a further recommendation at some future date as to its permanent adoption.’ The CinC went on to indicate that he was prepared to accept one complete flight of NCO pilots in each of the enlarged twenty-four-aircraft corps reconnaissance squadrons and up to 50% of all pilots in day bomber squadrons. He was less enthusiastic about NCOs flying fighters but, on a trial basis, he was prepared to have one flight of sergeants in one two-seater squadron and one flight in each of six single-seater squadrons. For night bombing duties, however, Haig considered that all pilots ‘must be officers.’ He imposed two other conditions. First, that the influx of NCOs was to be provided in a single group, not piecemeal, and, secondly, that separate messing facilities for sergeant pilots were to be provided on units associated with the trial.⁵⁶

With hindsight, and in the context of this book, the second of Haig’s conditions is of particular interest – more for what it implied than for what it actually said. While it was obviously impractical for sergeant pilots to share a mess with their officers, the idea that their introduction in relatively small numbers made it necessary to provide discrete accommodation for them was very odd, since every unit already had appropriate messing facilities for NCOs.

Was this curious decision the ultimate expression of the divisiveness which had been created by three years of treating pilots as an élite? After all, more than 200 noncommissioned back-seaters were already serving in France and no one had ever thought it necessary to make any special domestic arrangements for them. Yet the prospect of half-a-dozen sergeants turning up to fly Sopwith Camels prompted the immediate establishment of a mess for their exclusive use. To the paranoid, this decision could be interpreted as yet another manifestation of prejudice against back-seaters. Even to those who are not emotionally disturbed, however, it must surely be seen to illustrate the Army’s sadly blinkered approach to the realities of being an aviator. If three years of war had proved nothing else, it had surely shown that pilots and observers were two of a kind – two halves of a whole. It is arguable that neither of them needed to be segregated from their groundcrew colleagues, but if one did, surely both did.

London signalled its acceptance of the CinC’s constraints in November but, in view of the critical importance of short range tactical reconnaissance and artillery spotting in the eyes of the army-oriented RFC, the option of using sergeants for corps work was not taken up.⁵⁷ In effect, therefore, it had been agreed that the number of NCO day bomber pilots could be increased as required up to a maximum of 50% of establishment and that the use of a proportion of sergeants in selected fighter squadrons would be tried on an experimental basis. In practice, although a few sergeants did fly with bomber squadrons they never represented anything like half the overall strength and the aim of the trial began to focus increasingly on the single-seat fighter pilots.

Since there had been an embargo on the training of airmen pilots since the previous summer, this constraint would clearly have to be lifted and in February 1918 it was announced that training could be reinstated in batches of up to ten per month, although, in view of the manpower situation, ‘no skilled mechanics’ could be accepted.⁵⁸

Early ‘administrative difficulties’ were experienced in the training of NCO pilots and the programme soon began to slip. In January 1918 London notified GHQ BEF that the single-seat pilots were not now expected to become available until 15 March and that it would be the end of April before any two-seater pilots would be ready.⁵⁹

By then, however, the trial had been more clearly defined and it had been decided to concentrate solely on the single-seaters. The project now embraced twenty-four pilots who had been trained on SE5as and twelve trained on Camels; they were all now due to reach France on 25 March. In anticipation of their arrival HQ RFC issued instructions that Nos 1, 24, 41, 43, 60 and 70 Sqns were to prepare the necessary accommodation (two Nissen huts per unit) and arrangements were made to provide the additional domestic staff (one cook and one steward per unit).⁶⁰ This directive was promptly short-circuited by receipt of a letter from the Directorate of Training announcing that ‘the training of NCOs on other than 2-seater machines has not proved to be an unqualified success and there is no doubt that they have proved slow in taking to Scouts.’⁶¹ The upshot was that the trial was postponed indefinitely. It had not been abandoned, however, and the training staffs persevered throughout the summer.

In the event it would be September before the trial commenced. By then it had evidently been decided not to concentrate on single-seaters and only two fighter units (Nos 84 and 203 Sqns) were nominated to participate in the experiment. The other units involved flew two-seaters in the fighter reconnaissance role (Nos 11 and 48 Sqns) and as day bombers (Nos 103 and 206 Sqns).

In preparation for the arrival of the pilots involved in the trial, the orders concerning the provision of separate accommodation were reissued and, since two-seater units were now included, the instructions noted that any NCO observers on strength were also to be accommodated in the messes being provided for sergeant pilots – needless to say, no mention was made of the NCO observers flying with other squadrons.62

The new arrivals were initially to be held supernumerary to the nominal strength of each unit, being progressively absorbed against the establishment as replacements for officers who were lost or posted. As intended, most of the initial influx of sergeants did arrive in groups but these were supplemented by a gentle trickle of additional pilots to offset wastage. As a result, most units were exposed to more NCOs than the nominal six required by the trial. Apart from stipulating that these men were not to be transferred to other squadrons, HQ RAF deliberately laid down no policy as to how they should be employed. It was specifically left to the discretion of Brigade Commanders to decide whether to integrate them into their existing squadron organisations, or to concentrate them within all-NCO flights commanded by an officer.⁶³

Had the serious shortage of pilots anticipated in 1917 actually materialised there can be little doubt that the delays experienced in mounting the trial would not have been tolerated and that many sergeant pilots would have been sent to France, regardless of their capabilities. In practice the manpower problem had been solved, largely by the Dominions. An initial trickle of (mostly) South African and Australian cadets, arriving to be trained as pilots during 1917, had become a flood by mid-1918 and by that time substantial numbers of pilots were also beginning to graduate from the flying schools which the RFC had set up in Canada. When the output of the considerably expanded facilities in Egypt was added to the total, it seems possible that the number of officer pilots being turned out by the RAF’s global training organisation might even have begun to exceed its requirements by the summer of 1918 (much as it was to do in 1944 – see page 254-256).

Against this background, little real urgency appears to have been attached to the NCO trial and, the original dynamic underpinning it having evaporated, the main reason for sustaining the experiment was probably scientific curiosity. The suspense date for reports on the experience gained from the trial was 10 November 1918 and at least five of those submitted have survived.⁶⁴ Although the comments they contain are specifically concerned with pilots, there is much that can be read across to back-seaters so it is worth noting the gist of what was said.

There was almost universal agreement that the separate messes had been a serious mistake. It was considered essential that all pilots should share the same domestic facilities, partly because that was where flying was constantly discussed, allowing newcomers to soak up the experience of the older hands, and partly because it was where ésprit de corps was consolidated. Needless to say, no one actually made the point, but these arguments were equally applicable to back-seaters.

Opinions as to the overall capabilities of the sergeants involved in the trial varied considerably. For instance, Lt-Col T A E Cairnes, OC 22nd Wg, had considered that all of No 84 Sqn’s NCOs had been ‘good pilots’ and Brig-Gen C A H Longcroft, GOC III Bde, seems to have been equally content with those flying with No 11 Sqn who had ‘proved themselves to be quite as good as the average officer pilot.’ Brig-Gen Ludlow-Hewitt, GOC X Bde, was rather less enthusiastic about the experience of No 103 Sqn whose NCOs had been ‘thoroughly satisfactory when working in formations led by experienced Officers’ but lacking in ‘initiative and enterprise when flying alone on Reconnaissance or Photography.’

On the other hand, although No 206 Sqn’s style of operation will have been very similar to No 103 Sqn’s, its CO, Maj C T Maclaren, had considerable reservations about the performance of his sergeants, even on bombing raids, since ‘they do not keep good formation and when attacked are inclined to split up rather than packing together.’ While acknowledging that two of his sergeants had been satisfactory, in general he was of the opinion that ‘there is a marked difference between the NCO pilot and the Flying Officer, particularly with regard to reconnaissance and photographic work.’ Maclaren had tried his men with both commissioned and non-commissioned back-seaters and had concluded that the two-NCOs combination did not ‘possess the necessary intelligence and initiative for the carrying out of their work successfully’ and he had no doubt that ‘the absence of the spirit of the officer in command of the machine is largely felt.’ While the provision of a commissioned observer did improve matters, such mixed crews still tended to perform indifferently, because of a lack of mutual understanding between its members.

The most damning report came from OC 48 Sqn, Maj K R Park, who had received a total of nine sergeant pilots. He had had four of them posted home for further training and had recommended that a fifth be consigned to the trenches! Of the remainder, he considered only three to have been satisfactory. Park’s assessment led him to draw a very interesting conclusion. He was of the opinion that his three good NCOs had been the equal of officer pilots and he recommended that if a man is ‘of the right type and good enough to be a fighting pilot in a fighting unit, he should be commissioned.’ Much the same view had been reflected by both Cairnes and Maclaren. The latter, noting that his two satisfactory NCOs had both been educated at Public Schools, failed to ‘understand how they came out as NCOs as their flying is beyond reproach.’

With little else to go on, the RFC’s preference for commissioned pilots had always been based largely on instinct. The feedback from the formal attempt to assess the capabilities of NCO pilots had provided positive, if unscientific, evidence that the corps had been right to trust its judgement. While their assessments had been almost entirely subjective, three very experienced unit commanders had independently drawn the same conclusion, that a good pilot needed to possess much the same personal qualities as those traditionally associated with a commission. This tended to confirm the long-standing assumption that the terms officer and pilot were, in many respects, synonymous. The same was actually true of back-seaters, apart, perhaps, from those whose duties were confined solely to gunnery in those day bomber squadrons which usually operated in formation.

In closing this section, two further points should perhaps be made. First, while the trial had focused on a specific group of NCOs, others had still been reaching France via the usual channels, as they always had done, albeit in ever-decreasing numbers.65 When the fighting stopped in November 1918 only thirty-five of the 1,879 pilots on the strength of front-line units stationed in France, and operating under the control of HQ RAF, were NCOs⁶⁶ – less than 2% of the total. If the sergeants flying with units associated with the NCO trial are discounted, however, there were only two, or about 0.1%.

Fig 15. Casualties sustained during air-to-air engagements over France at selected periods.

Secondly, the Armistice had been signed just four days before the last report on the NCO trial had been rendered. Since the staffs soon became preoccupied with the problems associated with demobilisation, it seems unlikely that the, now largely irrelevant, reports will have attracted much attention. Nevertheless, anyone who did read them (and who was also able to read between the lines) might have been able to predict that the close correlation between commissions and, at least some, aircrew trades that had been highlighted by the trial might well present a difficulty in the future. The problem was that if a peacetime air force were to commission almost anyone who flew, most of them would actually have no one to command, which would make their being officers a little pointless and thus undermine the whole ethos of the commission. On the other hand, if it did not offer commissions on a fairly generous scale, it might not be able to attract sufficient volunteers of the necessary calibre. This peculiarly ‘air force’ problem was to crop up again and again over the rest of the century and, despite several attempts, it could be argued that it has never really been satisfactorily solved.

The ratio of casualties between Pilots and Observers.

Previous discussion of aerial gunnery, and of the observer’s role in air combat, raises the closely related question of casualties. There was (and still is) a widely held perception that observers tended to be more vulnerable than pilots, although this writer is not aware that any specific evidence has ever been offered to prove this. The publication in 1995 of Trevor Henshaw’s work on casualties has provided a readily accessible and reliable database, however, and this permits us to attempt some sort of analysis.⁶⁷ The results are presented at Figure 15, which examines selected incidents which occurred during three separate months in which air fighting was particularly intense.

Only the relatively personal nature of close quarter air-to-air engagements could have revealed any bias one way or the other, since ground fire was far too inaccurate to have been aimed at an individual crew member. For this reason casualties known, or likely, to have been caused by ‘Archie’ or by machine gun or small arms fire from the trenches have been excluded. Since fighting in the dark was not a significant factor either, incidents which occurred during night bombing missions have also been discounted, as have casualties sustained by two-seater pilots flying solo bombing sorties. It is acknowledged that the causes of some of the remaining incidents might be open to dispute but it is thought unlikely that there would be enough of these to alter the overall balance significantly. The tabulated figures are therefore believed to reflect, with tolerable accuracy, only those losses (killed or wounded in action or died of wounds) sustained as a result of air-to-air combat.

Since 55% of the sample were observers, it is clear that the observer/gunner of a two-seater operating in daylight was indeed statistically more likely to become a casualty than his companion. Why should this have been the case? One obvious explanation is that, since most attacks were delivered from a rear quarter, the attacker’s priority was to neutralise the return fire from the gunner. On the other hand there may merely have been a tendency to aim at the most obvious target, the occupant of the rear cockpit.⁶⁸ There is some evidence to support the second of these notions as it would explain the only significant anomaly which appears in the table, that of corps crews during ‘Bloody April’. At that time the majority of the aeroplanes being used in this role were still BE2s, in which the pilot sat in the back, and this appears to have been the only case in which he was more likely to become a casualty than his colleague.

A case in point. This DH 4 of No 5(N) Sqn crashed after having been shot up while attacking troops near Villers Bretonneux. Although the damage sustained is believed to have been inflicted by gunfire from the ground, rather than from a hostile aircraft, the injuries sustained by the crew conformed to the statistical pattern; the pilot, FSLt C J Haywood, was unhurt while the aerial gunlayer, T W Jones, was categorised as wounded in action. Interestingly, since this incident occurred on 30 March 1918, N5992 may have been the last RNAS aeroplane to be lost in action. (J M Bruce/G S Leslie collection)

There is another anomaly which should be pointed out, as it is not apparent from the tabulated figures, and this one tends to contradict the conclusion that he who occupied the back seat was invariably most at risk. The pilot of an FE2 also occupied the rear cockpit but in this case the ‘normal’ pattern of casualties was sustained, only (about) thirty-five pilots of the FE2s included within the April 1917 sample becoming casualties, compared to (about) forty-five front seat observers.

That might have been because the rear-mounted engine gave the pilot a degree of protection, or it could simply be that the inconsistency demonstrated by the FE2 is merely a result of the sample’s being too small to be statistically significant and thus too small to permit any definitive conclusions to be drawn. Indeed it could reasonably be argued that the samples used are too small to substantiate any worthwhile findings in every case, and that the whole argument therefore lacks rigour. Nevertheless, the overall pattern which emerges is remarkably uniform, indicating very similar percentile differences in almost every case, both over time and regardless of role. It is acknowledged that the foregoing may be a little superficial but, even so, its consistency does make it rather persuasive.

There is some anecdotal evidence to suggest that the inherent sympathy for pilots that was embedded within the system may also have exerted a negative influence on the survival rate of observers. For instance, writing some forty years after the event, one of No 5 Sqn’s Flight Commanders expressed his regret at the way in which he had persevered with an enthusiastic, but ham-handed, young pilot who was not really capable of handling an aeroplane and whom he should have sent home for further training.

This pilot was eventually killed in action, along with his observer, when his aircraft was shot down by an enemy fighter. To what extent his inadequacy as a pilot had contributed to the outcome of the engagement cannot, of course, be determined, but the fact that MRAF Sir John Slessor had ‘often kicked myself since for not having had the moral courage to get him taken off flying’ suggests that he certainly felt that it may well have been a factor. More to the point, however, Slessor also clearly felt some retrospective guilt over the fate of the unfortunate observer. As he put it, ‘It might have been different in a single-seater squadron, but we had the RE8 and I should have had more consideration for the poor devils who had to go up with him.’ 69

The overall statistics will also have been distorted by the many instances of pilots returning to base with a dead or wounded observer on board, whereas there were relatively few cases of this occurring in the reverse sense. This might appear to be a significant factor but it does not really alter the overall picture; it serves only to underline the disproportionate risk run by observers. As the Bailhache Committee had suggested in 1916, death through the incapacitation of their comrade-in-arms was a hazard virtually unique to back-seaters. For the purposes of this analysis, however, it is not really necessary to establish how each individual observer died, only to demonstrate that they did so in greater numbers than pilots.

________________________

¹      AIR1/31/15/1/156. Because of the considerable disturbance and a number of misunderstandings caused by the adoption of the new approach to pilot training, the Master General of Personnel, Maj-Gen G M Paine, summarised the arrangements and the impact that these would have on existing and future units and circulated this information in Air Ministry letter C.4519 dated 4 July 1918.

²      Despite the universal acceptance of Smith-Barry’s approach, the restructuring of the training system and the widespread availability of the Avro 504, problems persisted within the pilot training system, not least the provision of adequate numbers of instructors. Louis Strange (op cit, pp155-156) recalls that when he assumed command of 23rd (Training) Wing in March 1918 he found, among other difficulties, that he had only twenty instructors to handle 185 students. His solution was to ‘turn the next ten most promising pupils into instructors.’ In other words, to reinstate the discredited practice of 1915-16 (see Chapter 7, Note 11). Since the pupils of 1918 were being properly taught, however, ‘creaming off’ some of the more capable graduates was now a far more practical proposition. It certainly seemed to work, as Strange claims to have succeeded in significantly increasing the output of pilots while simultaneously reducing both the workload on the staff and the accident rate. It is sobering to note, however, that he also admits that, notwithstanding his improvements, aeroplanes were still being written off at a rate of one every 140 flying hours. In a unit the size of 23rd Wg (which at the time comprised No 81 Sqn and Nos 11, 39, 45, 60 and 61 Training Squadrons located variously at South Carlton and Scampton), this equated to ‘between thirty and forty machines a month, in addition to some seventy or eighty minor crashes.’

Unfortunately, while the confidence with which these fairly precise statistics are presented tends to persuade us that they must be accurate, they should be treated with some caution, as Strange may have been prone to exaggeration. For instance, he also claims, with equally convincing précis-ion, that his wing suffered sixteen fatalities in May 1918 alone. According to Chris Hobson’s authoritative Airmen Died in the Great War 1914-1918 (1995), the actual figure was eight.

³      Reference has already been made to the frequent reorganisation and redesignation to which all of the RFC’s training units were subjected but it may be of some interest to summarise here the progressive changes in nomenclature associated with the two longest-established observer training schools.

On 1 October 1917 the Wireless and Observers School moved from Brooklands to Hursley Park, Winchester, the dual functions of the school being separated shortly after its arrival. The training of wireless personnel was concentrated at the existing School for Wireless Operators at Farnborough, which reverted to its original designation of the Wireless School, RFC on 10 October. This title was to survive only for a month, however, as the division of responsibility was taken a stage further on 8 November when the unit at Farnborough became No 1 (Training) Wireless School, thenceforth dealing only with non-commissioned wireless mechanics and operators, No 2 Wireless School being set up at Penshurst at the same time to cater for W/T Officers, ie Equipment Officers specialising in wireless work. Now dealing solely with aircrew, the unit at Winchester needed to be renamed in any case, as its old title had failed to reflect the fact that it had been training large numbers of pilots for several months. On 11 October 1917 it became the Artillery and Infantry Co-operation School. Having moved to Worthy Down on 31 May 1918, it was renamed yet again on 19 September to become the RAF and Army Co-operation School.

Having taken up residence at Hythe on 27 November 1915, the RFC Machine Gun School became the School of Aerial Gunnery with effect from 13 September 1916. In January 1917 the unit was redesignated as No 1 (Auxiliary) School of Aerial Gunnery. On 9 March 1918 it merged with No 3 (Auxiliary) School of Aerial Gunnery (which had formed at New Romney on 1 August 1917) to form No 1 (Observers) School of Aerial Gunnery, both aerodromes remaining in use.

⁴      Largely as a result of the major upheaval within the training system, there had been an hiatus in the deployment of new two-seater squadrons. The last of the units formed under the old scheme to become operational had been No 110 Sqn which had crossed the Channel as long ago as the end of August 1918. None of the new-style squadrons was to see active service before the Armistice was signed but at that time the first of them, No 155 Sqn, was expected to take its DH 9As to France on 21 November with the Bristol Fighter-equipped No 138 Sqn to follow on the 30th. Neither went.

⁵      A consolidated strength return, reflecting the numbers and availability of pilots in all overseas theatres (other than France) had been routinely compiled on a weekly basis since mid-1916. Balloon observers were included from October but it was not until 1918 that anyone bothered to count the aeroplane observers on strength on a regular basis (although their numbers could be divined from individual squadron returns). The requirement for the numbers of effective and non-effective observers to be reported to London was contained in War Office telegram 49666 of 1 January 1918 (AIR1/398/15/231/39). The first return in which they actually featured was that rendered for the week ending 10 January by which time there were a total of 107 of them. The data in Figure 12 was taken from the return for the week ending 14 March which was dated the 15th and bore the reference E.4506 (AIR1/2364/226/7/3C).

⁶      On 31 October 1918, when the signing of an Armistice ended the fighting in the Near East, the RAF’s strength in qualified observers, including those who had been inherited with the ex-RNAS squadrons in the Aegean, amounted to 129 (of whom twelve were NCOs), plus seventy-six kite balloon officers. On that date there were a further ninety cadet/officer and five NCO observers under training in Egypt.

⁷      AIR1/17/15/1/87. Establishment 87/9406 dated 13 September 1917, which was specifically for a Service Squadron stationed in East Africa (which could, therefore, refer only to No 26 Sqn) included a note to the effect that, if necessary, NCO aerial gunners were to be substituted for officer observers. Paradoxically, no provision was made for any observers, of any rank. It is known, however, that numerous officers did fly on observation duties in that theatre, although this seems to have been done largely on an ad hoc basis. As an aside, it is interesting to note that the famous Wg Cdr L F W ‘Sos’ Cohen, who, at seventy years of age, added a DFC to his earlier DSO and MC by flying as an air gunner in WW II, did so while sporting an observers badge, worn on the strength of his having flown as such with No 26 Sqn in WW I. He certainly flew with No 26 Sqn, but there may be some doubt as to whether he actually satisfied all of the formal requirements entitling him to wear the badge.

⁸      While they were superficially similar, the flying training organisations established in Canada during the two world wars were fundamentally different. The WW I scheme was an entirely British undertaking, albeit lodged in Canada and provided with extensive facilities by the local administration. The constitution of the system set up during WW II became almost a mirror image of the original arrangement. Although it was given extensive material support by the British, especially in its early days, the Canadian element of the Empire Air Training Scheme of WW II was a Canadian responsibility from the outset.

⁹      It is appropriate to note here that, because Canada had not established an air force of her own, all Canadian (and British-trained American) aircrew flew with the RFC/RAF, back-seaters wearing the standard singlewinged flying ‘O’. By contrast, Australia had set up the Australian Flying Corps (AFC) and Military Order 801/1915 announced that any back-seaters qualifying (in Australia) after January 1916 would wear a badge with twin-wings flanking the letter ‘O’, the latter being embroidered in pale blue within a white laurel wreath. The paucity of photographic evidence suggests that this may not have been a very common emblem, however, and many Australian observers are known to have worn the British pattern badge, presumably because they had been trained in the UK, or because they were flying with RFC/RAF units.

¹⁰    For a comprehensive account of the history of the RFC in Canada see Dancing in the Sky by C W Hunt (Toronto, 2009).

¹¹    Prior to the introduction of role-specific establishments, deviations from the provisions of the standard establishment for an eighteen-aircraft Service Squadron dictated by the peculiar demands of each role had been authorised by appropriate endnotes, which were subject to fairly frequent amendment.

¹²    AIR1/1603/204/84/4. Amendment 87/RFC/1018 dated 22 December 1917. Apart from deleting the six NCOs from a corps squadron, this amendment also deleted six officer observers from an army squadron, leaving the latter with a notional fourteen officers (two with the HQ) and six NCOs. This rationalised the apparent oversight in the case of the army squadrons (which had never been planned to be expanded to twenty-four aircraft), although no remedial action seems to have been taken to rectify a similar, seemingly anomalous, increase in establishment which had been authorised for night flying squadrons (see Chapter 9, Note 45).

¹³    AIR1/17/15/1/87. Establishment for a twenty-four aircraft corps reconnaissance squadron, 87/9404 dated 19 January 1918.

¹⁴    AIR1/520/16/12/1. HQ BEF letter OB/1826E dated 5 March 1918.

¹⁵    Ibid. War Office letter 87/9406 (SD2) dated 19 March 1918.

¹⁶    AIR1/1163/204/5/2532. This file contains a headcount by squadron of pilots, observers and administrative and technical officers at an unspecified date, but clearly early in November.

¹⁷    AIR1/1135/204/5/2224. HQ 15th Wg letter GS.39 dated 15 March 1918.

¹⁸    Commanded by Maj A C Bishop, the RFC’s Armament School had originally been established at Perivale in June 1917 to provide training for technical personnel dealing with bombs, guns and synchronising gears. This unit was notionally disbanded on 18 December, its resources being used to establish a ‘new’ Armament School at Uxbridge on the same date.

¹⁹    The first two dozen (Arab-engined) Bristol Fighters earmarked for corps reconnaissance work were to have been sent to France in March 1918 when they were to be parcelled out on a six-aircraft per squadron basis, the rationale behind this being, in part, the implementation of the scheme to raise the overall unit establishment to twenty-four aircraft. Once that had been achieved, the remaining RE8s and FK8s were to have been replaced by Bristols through normal wastage until the equipment had been standardised. The supply situation did not permit much progress to be made with the re-equipment programme, however; so the numbers had to made up with the older types. The Bristols did not actually begin to materialise until June and experience soon showed that it would be better to allocate them to autonomous flights. Nevertheless, it was still the long-term intention to standardise on the Bristol Fighter and at the time of the Armistice current planning envisaged squadrons of RE8s and/or FK8s being re-equipped at a rate of one in November, one in December, five in January and eight in February. The remaining units, which were not to be re-equipped, were to ‘die out’ during March 1919.

²⁰    AIR1/1001/204/5/1260. Units flying the Bristol Fighter (and the earlier FE2 and 1 ½ Strutter) were (and still are) often identified as ‘two-seat fighter squadrons’. While this was a tolerably accurate description of their role it was not the officially approved terminology. The term ‘Army Fighter Squadron’ had been used in the summer of 1916 but AO 241 of 8 November 1916 proposed that the designations of all operational units should be standardised. HQ RFC agreed to the proposed new nomenclature which was formally introduced by AO 277 dated 21 December 1916. For the remainder of the war units operating two-seat fighters were always referred to as Fighter Reconnaissance Squadrons in official papers.

²¹    Louis Strange, op cit, p145.

²²    Frank T Courtney, op cit, p99.

²³    AIR1/920/204/5/885. The pilots engaged on this work included Capt A M Vaucour, Capt N Kemsley and 2/Lt T P H Bayetto. A summary of the conduct of the trial, including the conclusions drawn, was submitted to the War Office in Training Brigade letter TB/1067/2 of 26 April 1917. This was subsequently forwarded to HQ RFC who circulated its contents among its subordinate formations under letter CRFC 1708G dated 4 May 1917.

²⁴    Ibid. Report A/39 published by the RFC Experimental Station on 30 April 1917. A number of notable pilots had participated in the Orfordness programme, including Capts R H M S Saundby, E L Foot and H Meintjis and Lt J T B McCudden.

²⁵    Vivian Voss, Flying Minnows (1935), p106.

²⁶    The revised figure of 619 victories (and the apportioning of this total between pilots and observers/gunners) appeared in ‘Top Scorers – The Record of No 20 Sqn RFC/RAF’, an article by N Franks and F Bailey which was published in 1973 in Cross & Cockade (Great Britain) Journal, Vol 4, No 1. Subsequent research has indicated that No 20 Sqn’s total may actually have been as high as 630, this figure being published in Above the Trenches (1990), jointly compiled by C Shores, N Franks and R Guest.

²⁷    Arch Whitehouse, op cit, pp104-105.

²⁸    AIR1/400/15/231/41. Air Ministry letter 87/RFC/1106 (M5 A/122) dated 16 March 1918. Although it was not British policy to promote the cult of the ‘ace’, in the course of ratifying combat claims it was necessary to assign victories to individual pilots and it was inevitable that the identities of those who were particularly successful would become known and that their reputations would be established. In the two-seater world prior to March 1918, although combat reports (usually) make it clear whose gun had actually done the damage, and individual back-seaters will undoubtedly have kept their own tallies, officialdom automatically assigned victories to the pilot.

²⁹    AIR1/17/15/1/87. HQ RFC letter CRFC 1791(A) dated 6 December 1917.

³⁰    AIR1/1603/204/84/4. Amendment 87/RFC/1018, dated 22 December 1917, to Establishment 87/9406 of 19 May 1917. This change was amplified by War Office letter 87/RFC/1018(SD2) dated 31 December 1917 (AIR1/1078/204/5/1678).

³¹    Although these documents actually specified the requirements for back-seaters as officer and NCO observers, they still included footnotes indicating where it was acceptable to substitute gunners in the event that insufficient observers were available. An amendment was published on 20 February (AIR1/399/15/231/40), which deleted any further reference to aerial gunners from all unit establishments.

³²    AIR1/1214/204/5/2630. The data used to compile Figure 13 has been drawn from a nominal roll of all officers and non-commissioned aircrew serving with the RFC in France in January 1918. It is thought that this return probably reported the presence of all personnel who had been on strength at any time during the month but it is possible that it reflected the position on one particular (but unspecified) day.

³³    Ibid. 2/AM A Moult was with No 43 Sqn and 1/AM E R Perrett (7032) with No 60 Sqn.

³⁴    The arrival of six additional two-seater squadrons since March 1917 had been offset by the fact that six of the earlier units were now flying single-seaters, the reduction in the total available being accounted for by the loss of Nos 34 and 42 Sqns which had been transferred to the Italian Front in November 1917.

³⁵    These were Nos 58, 62, 83 and 216 Sqns while the previously noted No 69 Sqn had been redesignated as No 3 Sqn AFC.

³⁶    AIR1/1163/204/5/2532. Return of aircrew on strength by squadron dated 2 April 1918. Note that these totals do not include any of the ex-RNAS bomber units which had yet to respond to RFC-style reporting but it is likely that Nos 205, 206, 214 and 216 Sqns would, between them, have added another fifty or so back-seaters to the total.

³⁷    AIR1/400/15/231/41. Updating editions which had been effective since 29 January 1918, the following revised establishments (for units based in France) were published at the end of March 1918: AF/F/17 for corps reconnaissance squadrons, AF/F/18 for day bomber squadrons, AF/F/19 for night flying (ie FE2) squadrons and AF/F/20 for fighter reconnaissance squadrons.

AF/F/16, as reflected at Fig 14, covered RAF night bomber squadrons but it is worth noting that the RNAS, the original sponsor of the big Handley Pages, had its own views on the way in which they should be manned. By the spring of 1918 the establishment of a naval heavy bomber squadron provided for a crew of two pilots plus one gunlayer per aeroplane with five commissioned observers being held on the strength of the HQ.

³⁸    AIR1/1990/204/373/158. Establishment AF/F/46 of 16 July 1918 was drafted to cater for the specific requirements of HQ Independent Force. Note, however, that the two heavy bomber units which continued to operate under the aegis of HQ RAF (Nos 207 and 214 Sqns) were still supposed to be manned against AF/F/38 (which had superseded AF/F/16 on 23 May), ie twenty-three pilots and now only ten observers, all of them NCOs, although, in practice, both squadrons did actually have commissioned observers on strength (see Note 5 to Chapter 15).

³⁹    AIR1/1163/204/5/2532. HQ RAF return of numbers of aircrew by unit. Although undated, the squadrons listed and their assignment to wings and brigades fixes the date as being no earlier than 1 November 1918 and no later than the 9th.

⁴⁰    R Dallas Brett, History of British Aviation 1908-1914 (1933). When war was declared there were 305 military and 101 naval officers holding RAeC Certificates. By comparison only forty-seven non-commissioned army personnel and forty naval petty officers had qualified as pilots. Note that these figures do not represent an Order of Battle, since not all of these licence holders had subsequently been graded as military pilots while others had not maintained their currency.

⁴¹    WO123/55. A pamphlet, entitled Tests For First And Second Class Pilots Certificates For Officers And Men, was published with Army Orders for September 1913.

⁴²    AIR1/2148/209/3/191. Nominal roll of officers and non-commissioned aircrew serving with the BEF as at 15 September 1915. The two NCOs listed were Sgts F Courtney (2891) and W Watts (1831).

⁴³    AIR1/1290/204/11/71. Nominal roll of officers and non-commissioned aircrew serving with the RFC, dated April 1916. The three NCOs concerned were F/Sgt T Carlisle (351), Sgt T Bayetto (4808) and Sgt J Noakes (4469) respectively.

⁴⁴    Comprehensive records of the numbers of officer and NCO pilots trained during WW I do not appear to have survived but sufficient documentary evidence exists to provide a reasonable impression of the balance between them. Beginning in August 1912 (when the first batch was printed) the CFS supplemented the RAeC ‘ticket’ by issuing all pilots who successfully completed their subsequent training with an RFC Flying Certificate, those applicable to officers differing slightly from those provided for non-commissioned personnel, each series being numbered independently. Following the decoupling of military flying training from the RAeC in August 1916, the document issued to officers (but not that issued to mechanics) was superseded by a similar, but subtly different version known as a Graduation Certificate, although the original numbering sequences continued without interruption. It is known that Sgt E A Cook’s Flying Certificate No 175 was dated 7 April 1917 while 2/Lt H D Arkell’s Graduation Certificate No 5,321 was issued on 26 June. While these dates differ by several weeks, they are close enough to show that by the spring of 1917 officer pilots already outnumbered NCOs by more than thirty to one, and the gap will have widened even further during the remainder of the war.

⁴⁵    AIR2/9/87/7956. DAO letter 87/7956(AO1a) dated 25 June 1916. Subsequent correspondence amending the rules as initially published is on the same file.

⁴⁶    AIR2/9/87/7661. From the published (but less widely publicised) Appendix A to the Final Report of 17 November 1916.

⁴⁷    AIR1/1297/204/11/139. Nominal rolls of officers and non-commissioned aircrew serving with the RFC overseas, dated January-April 1917.

⁴⁸    AIR1/1297/204/11/140. Nominal roll of officers and non-commissioned aircrew serving with the RFC overseas, dated June 1917.

⁴⁹    AIR1/1301/204/11/158. Nominal roll of officers and non-commissioned aircrew serving with the RFC overseas, dated August 1917.

⁵⁰    Ibid.

⁵¹    AIR1/1036/204/5/1455. An enclosure on this file indicates that there were thirty-two NCO pilots flying with squadrons and a further seven held in reserve at Nos 1 and 2 Air Depots; unfortunately, this list is undated but there is circumstantial evidence to indicate that it was compiled circa late September 1917.

⁵²    AIR1/391/15/231/32. DAO letter 87/RFC/564 (O.2) dated 7 June 1917.

⁵³    AIR1/1302/204/11/167. Nominal roll of officers and non-commissioned aircrew serving with the RFC overseas, dated February 1918.

⁵⁴    AIR1/520/16/12/1. War Office letter 79/9962 dated 18 August 1917.

⁵⁵    Ibid. An appreciation derived by the Adjutant General at Advanced GHQ BEF from figures submitted by Lt-Col Festing in HQ RFC letter F.1958 dated 29 August 1917.

⁵⁶    AIR1/1078/204/5/1678. GHQ letter OB/1826/E/1 dated 6 September 1917.

⁵⁷    Ibid. War Office telegram 45677 dated 14 November 1917.

⁵⁸    AIR1/399/15/231/40. DAO letter 79/9962 (O.2) dated 15 February 1918.

⁵⁹    AIR1/398/15/231/39. DAO letter 79/9962 (O2) dated 8 January 1918.

⁶⁰    AIR1/1078/204/5/1678. HQ RFC letter CRFC 2022/1G dated 6 March 1918.

⁶¹    Ibid. Brig-Gen H D Briggs informed HQ RFC of the problems being experienced in training NCO pilots on single-seaters in Air Ministry letter 79/9962 dated 5 March 1918.

⁶²    Ibid. HQ RAF letter 2567(A) dated 30 August 1918.

⁶³    AIR1/1078/204/5/1678. HQ RAF letter 2567(A) dated 30 August 1918.

⁶⁴    AIR1/1036/204/5/1455. Reports to HQ RAF, submitted either directly or via Brigade HQs, representing the views of Nos 11, 48, 84, 103 and 206 Sqns are on file.

⁶⁵    For example: Sgt J Matthews and FSgt J Helingoe had arrived in France with No 148 Sqn in April 1918 as had Sgt H N Lee of No 88 Sqn; Sgt H H Wilson joined No 99 Sqn in June and at the end of August, Sgts H W Tozer and A Haigh reached France with No 110 Sqn; Sgt A O Hall was posted to No 70 Sqn on 5 September 1918 and, just as hostilities were ceasing, the redoubtable FSgt Ernest Elton was on his way back to France with No 39 Sqn. This list is not presented as being exclusive; there will probably have been (a few) others.

⁶⁶    AIR1/1163/204/5/2532. HQ RAF return of numbers of aircrew by unit. Although undated, the assignment of squadrons to wings and brigades fixes the date as being no earlier than 1 November and no later than the 9th.

⁶⁷    Trevor Henshaw, op cit.

⁶⁸    At the close ranges and low speeds at which air combat was conducted during the biplane era, it was perfectly feasible to aim with such precision. This was specifically required by the post-war aerial firing test for air gunners, first promulgated in AMWO 624 of 4 August 1921 and later incorporated into both KRs&ACIs and AP1112. The test required a gunner’s camera gun photographs ‘to show the opposing pilot not more than ¼ inch from the correct point of deflection.’ It was not enough simply to spray the other aeroplane with bullets; the aim was to ‘shoot to kill’.

⁶⁹    In his These Remain, p71, Slessor identifies the crew as Porter and Morris. They were actually Lt Edward Pember and 2AM Arthur Morley who were shot down on 30 September 1917 by Vizefeldwebel Ernst Hamster flying an Albatros of Ernst Udet’s Jasta 37. (See also Note 25 to Chapter 6).