Gran’s changing accounts of a Scott–Oates argument in February 1911

One anecdote from Gran’s memoirs relates an argument between Scott and Oates in February 1911, during the depot journey. At this point, Oates, in charge of ponies, wanted to kill weak animals prematurely, whereas Scott wanted every pony preserved for the November–December 1911 polar journey transport. Attempting to preserve the weakest ponies, Scott ordered the depot to be laid early and for everyone to return to base.

Here, Gran attributes dialogue to Scott and Oates: however, Gran’s story keeps changing across his memoirs:

1. (1) 1915, Hvor sydlyset flammer [Where the Southern Lights blaze]: There is no mention of any argument (Gran, Reference Gran1915, p. 42; Gran, 1915/Reference Gran and Hattersley-Smith1984, pp. 58–59).

2. (2) 1924, En helt [A hero]: The first version of the Scott–Oates argument:

   [Oates] dared to say to his boss on this occasion, when he heard his decision: “You will regret it, sir.” “Regret it, my dear Oates, or not,” Scott replied earnestly, “I am acting as a Christian.” [[V]ovet [Oates] sig til ved denne anledning der ute paa barieren, at si til sin chef da han hørte hans beslutning: «De vil angre det, min herre.» «Angre det, min kjære Oates, eller ikke», svarte Scott alvorsfuldt – «jeg handler iafald som et kristent menneske.»] (Gran, Reference Gran1924, p. 54)

3. (3) 1945, Slik var det [So it was]: Gran repeats the 1924 Scott–Oates dialogue (Gran, Reference Gran1945, p. 107).

4. (4) 1961, Kampen om sydpolen: The second version of the Scott–Oates argument:

   [Oates] said to Scott that he considered it a big mistake not to continue. “Wrong or not,” Scott replied, “I’ve had more than enough of this cruelty to animals, and for the sake of some days’ progress I will not betray my feelings.” [[Oates] sa rett ut til Scott at han anså det for et stort feilgrep ikke å fortsette. «Feilgrep eller ikke», svarte Scott, «jeg har hatt mer enn nok av dette dyrplagieret, og for noen få dagsmarsjers skyld vil jeg ikke trosser mine følelser.» (Gran, Reference Gran1961, p. 106)

5. (5) 1974, Fra tjuagutt til sydpolfarer [From kid to South Pole explorer]: Gran reverts to the 1924 version of the Scott–Oates dialogue (Gran, Reference Gran1974, p. 240).

Gran’s Scott–Oates dialogue has been repeated by later secondary historians as “evidence” that Scott spared the ponies due to “Christian” sentimentality. However, Gran’s later-life versions are the earliest known source for this anecdote. The Scott–Oates dialogue changes back and forth with Gran’s successive memoirs: a credible eyewitness account would give a single version of this argument, not two. One later writer, noting the discrepancy, attempts to resolve the contradiction by splicing together both variants to form a single longer Scott–Oates conversation (Smith, Reference Smith2002, p. 143). However, the existing evidence indicates that Gran never transcribed a longer conversation during 1910–12. These alternating variants appear to be Gran’s post-1924 experiments in creating dialogue.

Why would Gran create dialogue for Scott and Oates? One should understand that Gran had no university-level training in the humanities; also, in writing for a popular market, he probably believed some illustrative embellishment was permissible. Since Atkinson claimed during 1912–13 that Scott had given “strict injunctions that the dogs should not be risked”, it would have seemed a logical extrapolation to Gran by 1924 that Scott’s wish to preserve animals in February 1911 was due to sentimentality.

We doubt Gran intended to mislead; we believe that he intended to illustrate, with hypothetical dialogue, his understanding of the “characters” of Scott and Oates, and, by so doing, “solve” the mystery of the prematurely curtailed depot journey. However, the danger of “attributed dialogue” in later-life memoirs is its obscuring of more reliable contemporary testimony, such as Scott’s own journal.

In February 1911, Oates effectively offered Scott a choice between killing poor-quality ponies or returning to base prematurely (in an effort to keep animals alive for longer). Clues to Scott’s thought processes in 1911 exist in Scott’s contemporaneous journal: these do not indicate a motivation of sentimentality.

Scott’s 1911 journal reveals Scott’s conviction that every animal should survive to haul loads on the November polar journey and acknowledges the eventual “sacrifice of animal life”: not sentimentality, but pragmatism. It also reveals that Scott believed Oates should never have brought weak animals on the depot journey: Oates’ plan to kill weak animals in February 1911 conflicted with Scott’s wish to preserve all animals until the November–December 1911 polar journey. During that polar journey, every surviving pony was employed in load hauling, to “get better value out of them” and then ultimately killed for meat to feed dogs and men. Hence, if a Scott–Oates argument had occurred in February 1911, we doubt that it followed Gran’s later-life attempts at character exposition through dialogue. If Scott was frustrated in February 1911, this was probably due to Oates’ selecting poor-quality ponies for a depot journey and not respecting Scott’s wish to preserve every animal until November.

We believe that the motivation of “sentimentality” attributed to Scott in February 1911 is Gran’s later-life hypothesis, based on hindsight and independent extrapolation from other expeditioners’ memoirs. Gran’s notion of Scott as supposedly placing priority upon animal longevity is contradicted by the immediately relevant evidence in Scott’s 1911 journal showing Scott evaluating animals’ utility and acknowledging the “sacrifice of animal life”. Scott’s own testimony on his motivation in 1911 therefore contradicts Gran’s later-life statements.

Gran’s mere presence on the expedition, even as a junior expeditioner with imperfect English who was absent from many crucial events, has unfortunately resulted in certain historians taking all Gran’s writings as unimpeachable eyewitness evidence. Efforts have not always been made to separate Gran’s more reliable contemporary record from his later speculative hypotheses on events he did not witness or may have misunderstood or misremembered. When speculating on Scott’s thought processes to explain his actions, it is safest to be guided by Scott’s contemporary journal rather than an observer writing decades later with alternating narratives.

Distinguishing reliable sources and identifying hypotheses

When faced with diverging narratives of an event, historians must separate reliable written contemporary documentation from uncorroborated later-life rumour. In George Lewis’s words, “History is not a matter of opinion, but a matter of sources”. Historians need to evaluate evidence for bias and external influence and reject unreliable “outliers”. Gran, in Race for the South Pole, relates versions of events that contradict contemporary accounts: since Gran’s versions originate from a non-eyewitness source, first appear decades after the fact, and are uncorroborated by contemporaneous eyewitness sources, they are safest regarded as “outliers”.

Unfortunately, when Huntford first repeated these stories in 1979 (the “January 1912 verbal orders to Evans”; the Campbell–Amundsen conversation), he cited no source. Polar researchers, unable to evaluate their veracity, accepted them as cast iron fact. Now, Huntford has confirmed Gran as his source, one must ask why Gran related these anecdotes. We believe that, in writing his later-life narratives, Gran shifted from his expected role of “eyewitness” into a more evaluative role, that of secondary historian. By 1945, when Evans became “Mountevans”, Scott’s journals (1913), Evans’ South with Scott (1921) and Cherry-Garrard’s The Worst Journey in the World (1922) had seen publication: all narrated details unknown to Gran during his time on the expedition.

Given fresh information, Gran faced the conundrum of reconciling two contradictory sources pertaining to the cause of the Polar Party’s deaths. The first was Atkinson’s (Reference Atkinson and Scott1913) claim, repeated in Cherry-Garrard’s (Reference Cherry-Garrard1922) memoir, that by February 1912 Scott had given “strict injunctions” that “the dogs should not be risked” (Atkinson, Reference Atkinson and Scott1913; Cherry-Garrard, Reference Cherry-Garrard1922). The second was Scott’s journal entry expressing surprise that dog teams had not reached Mount Hooper depot by 10 March 1912: “the dogs which would have been our salvation have evidently failed” (Scott, Reference Scott and Jones2006, p. 408). How could Gran reconcile Atkinson’s claim in early 1912 that Scott had ordered that “the dogs should not be risked” with Scott’s recorded disappointment in March 1912 that dog teams had not arrived?

One means of reconciling both sources would be to construct an overarching hypothesis, a narrative describing a possible scenario which would allow both testimonies to be true. One analogy for this process is “detective work”, where evidence at the “crime scene” is incorporated into a hypothetical chronological reconstruction of the crime that provides a theoretical sequence of events and makes sense of the available evidence. It appears that Gran believed both Scott’s and Atkinson’s statements were accurate and required inclusion, so in Race for the South Pole, Gran created the following hypothetical events to resolve the contradiction:

1. (1) Extra “verbal orders for the dog-teams”, supposedly given by Scott to Evans on 4 January 1912;

2. (2) Evans’ supposedly failing to deliver Scott’s “verbal orders” at base in February 1912.

Gran did not create this hypothesis ex nihilo: his 1961 memoir demonstrates his sincere misunderstanding of Scott’s 1911 orders (his erroneous belief that Scott delayed giving orders for dog teams until the last minute). By building on this misunderstanding to hypothesise that Scott gave “missing last-minute orders to Evans” in January 1912, Gran’s narrative resolves the discrepancies. It hypothesises that Scott cancelled his orders verbally in December 1911 (explaining Atkinson’s statement) and then afterwards re-instated his orders to Evans in January 1912 (explaining Scott’s later disappointment), and hypothesises that Evans’ near-fatal scurvy prevented Scott’s orders being delivered to base. By creating additional events to construct a hypothesis, Gran’s narrative presents both Scott and Atkinson as speaking the truth.

However, when compared against contemporary testimony, this hypothesis is unsatisfactory. The problem is that the claim of Scott’s “verbal orders” apparently originates with Atkinson, not Scott himself: we must compare Atkinson’s 1912–13 claims with Scott’s 1911–12 recorded behaviour. Scott’s expedition protocol, in accordance with naval custom, was to issue changes of order with written instructions. When sending back the Motor Party on 24 November 1911 with updated orders to send out a man-hauling team to transport three of the originally specified 5 X.S. rations to One Ton Depot, Scott addressed this written update to George C. Simpson, the de facto head of base. Simpson transcribed Scott’s written orders into his own handwritten 1911–12 journal:

It should be noted that this document, delivered to Simpson by Bernard Day, does not constitute a cancellation of the dog teams: while dogs “may” be non-existent (due to hazardous Antarctic conditions), this is Scott’s warning of possible difficulties, not his cancellation. The existence of this transcription in Simpson’s 1911–12 journal establishes Scott’s custom for updates as sending written instructions for Simpson’s attention. Had either Atkinson or the dog driver Meares been given updated orders from Scott during 1911 to deliver to base, these would have been written, and Simpson would have recorded Scott’s written updates verbatim within his journal. However, no further updates from Scott were recorded by Simpson in his 1911–12 journals.

The situation grows more complex when one understands that the dog driver Meares engaged in suspicious-looking behaviour in January 1912. One should note that Meares’ Reference Meares1918 Air Force service record describes him as “Expe[rience]d in Exploration and Intelligence Work” (Meares, Reference Meares1918), and as having participated in the “Russ[ian]-Jap[anese] War” [1904–5], which could have only been in an intelligence capacity. This previous intelligence experience explains Meares’ swift promotion and appointment to “Intelligence and Transport Officer” for the Royal Naval Air Service in France in 1915 (May & Lewis, Reference May and Lewis2015, p. 662). Meares’ service record by 1910–12 establishes him as a British intelligence agent and therefore familiar with the practices of dissimulation and spreading disinformation.

Evidence exists of Meares’ creation of false narratives to evade accountability. For example, Meares claimed in a 1910 newspaper interview that he personally purchased Scott’s expedition ponies in Siberia: he “knew the types [of pony] well”. However, by 1911, when the ponies’ poor condition became obvious, Meares subsequently spread contradictory false narratives concerning the purchase. To Wilson, Meares claimed that “a dealer and a vet” in “Vladivostok” [Siberia] had been misleading; to Frank Debenham, that a nameless “friend” of Meares had purchased them in “Mukden” [Manchuria] and that he, Meares, was “no judge of horse flesh”. This claim of equine ignorance is particularly unconvincing, as Meares had been a non-commissioned cavalry officer in the Boer War during 1901–1902. A study of Meares’ shifting narratives concerning the ponies’ purchase, and evident disinformation in a 1910 Reuters report of his travels, was published in 2015 (May & Lewis, Reference May and Lewis2015).

On 17 January 1912, Meares was recorded by Simpson at Cape Evans as ready to depart for One Ton depot, with loaded sledges, when the Terra Nova was unexpectedly seen at sea. This is the account in Simpson’s 1911–12 holograph [handwritten] journal:

Simpson testifies that on 17 January Meares claimed that his original intention had supposedly been to transport two sledges containing “luxuries” from Cape Evans to One Ton depot (in Chief Stoker William Lashly’s January 1912 table of polar journey distances, a journey of 132.5 geographical miles [245 km] [Lashly, Reference Lashly and Ellis1969, p. 141]). This story, that Meares’ sledges supposedly contained “a stock of luxuries like Irish Stew, Marmalade and Tinned Fruits”, arouses suspicion immediately. A dog driver does not embark upon a 490-km round trip in Antarctica merely to deliver “luxuries”: this would waste time and the dogs’ strength. Indeed, in his October 1911 written orders, Scott emphasised that the dogs should not be exhausted by inessential tasks: Scott specified that if Meares had completed his “second journey” in good time and were later requested to unload the ship before the “third journey”, “I should like you to give such assistance as you can without tiring the dogs” (Evans, Reference Evans1921, italics ours).

Since Scott had ordered Meares not to “tire the dogs” with inessential duties and had left no written orders for delivering “luxuries”, and since Meares would only plan a 490-km round trip with dog teams in Antarctica were it strictly necessary, the most realistic scenario for Meares’ loaded sledges and initial readiness to leave for One Ton on 17 January 1912 is that Meares’ sledge cargo contained not “luxuries”, as claimed to Simpson, but necessities for the “second journey”, which had been specified in Scott’s 1911 written orders (i.e. delivering the remaining 2 X.S. rations, fuel and “as much dog-food as you can carry” to “One Ton Depot” in January 1912). Delivery of that “dog-food” to One Ton in January would enable the dog teams to progress beyond One Ton on their “third journey”, beginning in February, and consequently intercept the polar party further south in March 1912 (an outcome specified in Scott’s 1911 orders). The correlation of “One Ton Camp” in Simpson’s 1911–12 journal with “One Ton depot” in Scott’s October 1911 orders indicates that the “second journey” was the task Meares originally intended to fulfil on 17 January 1912.

So why did Meares claim he was only transporting “luxuries” to One Ton on 17 January? It is most plausible that Meares misrepresented his sledge cargo as inessential “luxuries” to Simpson to avoid having to leave Cape Evans at a time when the ship might arrive and depart without him. In recording Meares’ account in 1912, Simpson evidently trusted his word: as Simpson had not participated in the January–February 1911 “depot journey” to build One Ton, he would not have understood the distances involved (or the sheer implausibility of Meares’ claim to transport “luxuries” so far). As Scott had earlier stated that updates would be sent back with “returning parties” (Evans, Reference Evans1921), and as Simpson was a civilian scientist, not a naval officer, Simpson would not have detected the incongruity of Meares’ claimed alteration of Scott’s expectations being belated, and verbal rather than written.

Meares had an evident personal wish to leave Antarctica before the expedition’s official end in 1913. In October 1911, Oates recorded in a letter that Meares already intended to “clear out whatever happened” (Oates, Reference Oates1911); the next chance to leave Antarctica was known to be early 1912, when the ship would arrive at Cape Evans. Given this context of Meares’ declared intention to leave Antarctica prematurely, it appears that on 17 January 1912 Meares consequently spread disinformation, misrepresenting his cargo as “luxuries” and his journey as easily cancellable.

The evidence from Simpson’s 1911–12 journal, and this hypothesis concerning Meares’ behaviour, was presented in a 2014 Polar Record article (May & Airriess, Reference May and Airriess2014) and restated in 2018 (May, Reference May2018a). Meares never explicitly claimed in January 1912 that Scott had cancelled/curtailed his dog teams; instead, Meares’ implication was that everything necessary was already at One Ton depot and his cargo mere “luxuries”, so Scott needed no further action taken. However, this was untrue. Meares’ refusal to depart for One Ton resulted in an absence of dog food at One Ton, which would limit the dog teams’ later “third journey”. Had Scott ever ordered Meares to curtail/cancel the dog teams’ progress and not undertake the “second journey”, such orders would have been written and appear in Simpson’s journal: no such orders exist there. Seen objectively, Meares’ behaviour on 17 January 1912 smacks of dishonest, last-minute improvisation.

As May stated in 2012, it was implausible to suggest that Scott had issued orders protecting the dogs: Scott would not have curtailed/cancelled a journey which would have served as a “safety-net” for his Polar Party (May, Reference May2012). In 1910–11, Scott brought animals to Antarctica for transport, fully intending “the sacrifice of animal life” if necessary (Scott, Reference Scott and Jones2006, p. 116). Scott already knew that men could suffer illness or injury during an expedition: during the Discovery expedition (1901–04), Scott’s “Farthest South” journey had foundered when one team member, Lieutenant Ernest Shackleton, became debilitated with scurvy. Their subsequent desperate march back to base taught Scott the importance of a “safety-net” in Antarctica. Consequently, should Scott’s companions suffer debilitation during the 1911–12 polar journey, Scott’s 1911 written orders would summon dog teams for support during March 1912. Having written this expectation into his October 1911 orders, it is implausible that Scott would destroy this “safety-net” during his November–December 1911 trek.

It should be understood that whenever Scott made an error during the expedition, he would record it in his journal with harsh self-criticism. On 8 January 1911, a motor sledge sank irretrievably through ice in McMurdo Sound. Scott’s journal contains self-reproach: “I stupidly gave permission for the third motor to be got out this morning” (Scott, Reference Scott and Jones2006, p. 81). On 14 January, Scott comments of a pony’s death, “I blame myself for not supervising these matters more effectively” (p. 124). On 18 March 1912, Scott blamed himself for his frostbitten foot, which hampered the Polar Party’s progress:

Phrases such as “I stupidly gave permission”, “I blame myself” and “like an ass” indicate that Scott identified and recorded his errors without self-excuse. Therefore, if Scott had made the misjudgement ascribed to him by Gran’s hypothesis (cancelling/curtailing the dog teams, then belatedly sending back fresh orders with Evans in 1912), then we would expect Scott’s March 1912 journal to contain speculation and self-criticism explicitly linking the dog teams’ failure to arrive with his earlier cancellation/curtailment. However, no mention exists of any mid-expedition cancellation/curtailment of dog teams within Scott’s 1911–12 writings: it appears that Scott always expected Meares to restock the depot with dog food (the “second journey” of January 1912) ready for the “third journey” of February–March 1912. Furthermore, the Polar Party’s decision to support Oates’ slowed progress between 2 and 16 March 1912, and Oates’ decision (after disclosing frostbite) to continue for another 14 days before his self-sacrifice indicates the Polar Party’s unchanged belief that dog teams would meet them in time (May, Reference May2012).

When Huntford in 1979 related two previously unheard anecdotes (“Scott’s verbal orders to Evans”, “Amundsen’s fear of British motor-sledges”), some writers repeated them as fact. In 2012, May analysed the contradiction between Atkinson’s (Reference Atkinson and Scott1913) claim and Scott’s (1912) surprise at the dogs’ failure to arrive at Mount Hooper. From the available evidence, it appeared to May that Huntford in 1979 had created the hypothesis of last-minute verbal orders to resolve this contradiction: “Huntford’s scenario of Scott’s last minute verbal ‘change of orders’ on the polar plateau in January 1912 tied up these inconvenient loose ends very neatly indeed, but unfortunately Huntford’s scenario was pure invention based on an error. It has led a number of polar historians down a regrettable false trail” (May, Reference May2012, p. 88).

Now the discovery of Gran’s Race for the South Pole indicates that the narrative of the “Scott’s last-minute verbal orders to Evans” was actually Gran’s (1945/post-Reference Gran1945) hypothesis. We now understand that Huntford’s error was not in original creation, but in not crediting Gran for the story. Had Huntford credited Gran in 1979, it would have been clear that Gran had originated the “pure invention based on an error”. Hypotheses are external constructions, designed to explain the facts: historians must prevent their being confused with extant facts. In 1979, Gran’s statements should have been identified as later-life hypotheses, then checked against primary evidence for reliability. However, it appears these checks were not made. Instead, Gran’s hypotheses were repeated as cast iron fact, distorting understanding of the expedition for decades.

In the next section, we present fresh evidence tracing Atkinson’s changing story between 1911 and 1919, and why we believe Atkinson accused Meares of having “disobeyed [Scott’s] orders.”

Atkinson’s testimonies: 1911, 1912–13, 1919

In 2012, May presented her hypothesis, following the available evidence, that Atkinson had misrepresented Scott’s orders due to an unforeseen crisis at base (Evans’ near-fatal scurvy). In 1912, Atkinson had given Cherry-Garrard orders that Scott wished for dogs to be protected, and no instructions for Cherry-Garrard to proceed past One Ton depot. Consequently, during his southern journey in February–March 1912, Cherry-Garrard remained at One Ton instead of slaughtering spare dogs to provide meat to progress further. In 1913, Atkinson wrote that “Strict injunctions had been given by Captain Scott that the dogs should not be risked in any way” (Atkinson, Reference Atkinson and Scott1913, p. 669). Repeated by Cherry-Garrard in 1922, this apparently influenced Gran’s later-life hypothesis that Scott’s motivation was animal sentimentality.

However, what Gran in Norway could not have known is that Atkinson later privately repudiated this story of Scott’s “strict injunctions”, in a 1919 letter to Cherry-Garrard discussing Meares:

Public accusation of Meares was not possible in 1919: neither Atkinson nor Cherry-Garrard possessed Scott’s October 1911 orders (“no proof in writing”). However, Scott’s orders presently exist in two corroborating transcriptions (Evans, Reference Evans1921; Wilson, Reference Wilson1911). The 2014 May-Airriess article details Scott’s orders for the dog teams and identifies Meares’ refusal to restock One Ton with dog food on 17 January 1912 (neglecting the “second journey”) as the specific instance of Meares’ failure to follow Scott’s orders (May & Airriess, Reference May and Airriess2014).

Further evidence exists that Scott never ordered Atkinson in 1911 to protect dogs. In 2017, Gran’s handwritten 1912 journal was purchased by Canterbury Museum (New Zealand): inside, Gran transcribed Atkinson’s journal entries for 29 October–27 December 1911. As Atkinson’s journal is currently lost, this transcription is the sole surviving record of Atkinson’s December 1911 understanding. In this transcription, Atkinson never mentions Scott’s ordering any cancellation/curtailment:

Had Scott in 1911 given Atkinson new orders involving curtailment of the dog teams, one would expect Atkinson to record this in his journal: Atkinson did not, nor did he, upon his return to base, deliver any update from Scott for Simpson’s transcription.

Furthermore, the fact that Atkinson accused Meares in 1919 confirms that Scott never gave Atkinson such “injunctions”. Had Scott instructed Atkinson in 1911 to protect dogs, then Atkinson would know beyond question that these were Scott’s wishes: there would therefore be no reason for Atkinson later to accuse Meares of “disobey[ing] orders” in failing to restock One Ton. The fact Atkinson accuses Meares in 1919 indicates that Scott never gave Atkinson these “injunctions” in 1911. (Indeed, in 1913, Atkinson uses the passive tense, “Strict injunctions had been given”, without indicating the recipient: had Scott told Atkinson this, one would expect Atkinson to state this directly.)

In 2012, May’s article presented the hypothesis that Atkinson had independently invented these “strict injunctions” to prevent Cherry-Garrard from travelling beyond One Ton on the Great Ice Barrier, since Cherry-Garrard (myopic, nervous, inexperienced) could have endangered himself and companion Dmitri Gerof by getting lost on the less well-marked route beyond One Ton. Indeed, Cherry-Garrard during his return from One Ton left a note dated 16 March 1912 at a later depot, admitting “We have made no depôts on the way in being off the course all the way” (Cherry-Garrard, Reference Cherry-Garrard1912). From 10 to 16 March 1912, Cherry-Garrard evidently lost his way along the well-marked route north from One Ton back to Hut Point, a fact that indicates possible catastrophe had Cherry-Garrard attempted to progress south from One Ton along a less well-marked route.

Cherry-Garrard, by his own account, was nervous and unskilled in dog driving and navigation in February 1912: “I had never driven one dog, let alone a team of them; I knew nothing of navigation; and One Ton was a hundred and thirty miles away, out in the middle of the Barrier and away from landmarks” (Cherry-Garrard, Reference Cherry-Garrard1922, p. 417). Atkinson unquestionably took charge at Hut Point during this period, since Evans was incapacitated by near-fatal scurvy (“it was decided by Atkinson that I was to take out the dogs”; “my orders were given me by Atkinson” (pp. 415, 416)). In February 1912, Atkinson gave Cherry-Garrard no orders to travel beyond One Ton, in contravention of Scott’s 1911 written expectations that dog teams should travel south of this depot as far as 82′30″°S. Atkinson was the earliest known source for the unlikely statements that Scott gave “strict injunctions” protecting dogs and that “Scott was not in any way dependent on the dogs for his return” (p. 416). Since Cherry-Garrard was evidently unprepared for the task assigned him, May hypothesised in 2012 that Atkinson had noted this and improvised the story of Scott’s “injunctions”, limiting progress to One Ton to prevent Cherry-Garrard from fatally endangering himself and Gerof (May, Reference May2012).

However, in 2012, May was unaware of the presence of an intelligence agent, Cecil Meares, on the expedition, and of Meares’ habit of relating false stories to evade accountability. Now Meares’ habit of false narratives has been identified (May, Reference May2018a; May & Airriess, Reference May and Airriess2014; May & Lewis, Reference May and Lewis2015), and we have revised our hypothesis of Atkinson’s role in the tragedy as follows:

1. (1) Scott never cancelled his original October 1911 expectation for dog teams to meet the Polar Party out on the Great Ice Barrier, south of One Ton depot, by March 1912. Neither Meares nor Atkinson delivered any written cancellation for Simpson to transcribe in his 1911–12 journals.

2. (2) Given this, Atkinson’s unsubstantiated statement that “Strict injunctions had been given by Scott that the dogs should not be risked in any way” was anomalous. Atkinson had either invented this independently or heard it from a third party. In 2012, in the absence of evidence suggesting otherwise, May hypothesised that Atkinson had created this narrative to protect Cherry-Garrard.

3. (3) The existence of Gran’s (Reference Gran1912) transcription of Atkinson’s 1911 journal, containing no orders from Scott cancelling/curtailing the dogs, indicates that Scott gave Atkinson no such orders in 1911 (Gran, Reference Gran1912, p. 235). However, May’s 2012 hypothesis that Atkinson independently created this story of “Scott’s strict injunctions” in 1912 requires revision. Considering Meares’ evident skill at creating false narratives (previous implausible/contradictory stories to Wilson, Debenham, Simpson and Reuters), it now seems most plausible that Atkinson received this story of “Scott’s strict injunctions” from Meares, the man responsible for the dog teams, and that Atkinson later repeated Meares’ story in good faith. Faced with the unstocked depot and a claim that Scott required dogs not to be “risked”, Atkinson in 1912 would have trusted Meares’ word.

4. (4) In 1919, Atkinson accused Meares of “disobey[ing] orders”. We hypothesise that in January–February 1912 Meares deflected others’ attention from his failure to fulfil the “second journey” in the same manner as his earlier purchase of poor-quality ponies: disseminating verbal disinformation to his colleagues to evade responsibility.

What triggered Atkinson’s belated realisation of Meares’ “disobedience”? In the British Library, we located Atkinson’s personal two-volume 1914 edition of Scott’s Last Expedition, inscribed “E.L. Atkinson, Bois d’Ohlain, 11 May 1917” [shelfmark X.800/30141]. Close examination of both volumes reveals no page annotations; however, this inscription indicates that Atkinson re-read Scott’s journals in France in 1917. We believe that, after many years, Atkinson could finally read Scott’s journals with sufficient detachment to understand that Scott, in March 1912, fully expected the dog teams to reach Mount Hooper depot (61 geographical miles (112 km) south of One Ton) and must, therefore, have expected One Ton to be stocked with dog food. Scott’s March 1912 expectation, therefore, indicates that One Ton’s lack of dog food in January 1912 was Meares’ independent omission, not Scott’s instruction. We believe that Atkinson’s 1917 re-reading clarified his understanding of Scott’s intentions, prompting his 1919 conclusion that Meares had “disobeyed orders”. Atkinson’s behaviour fits that of someone who never received strict injunctions from Scott in 1911, but instead heard this story from Meares at base in February 1912 and repeated this as truth until later re-assessment caused him to accuse Meares of “disobey[ing] orders”.

Cherry-Garrard wrote in 1922 that, by February 1912, One Ton had not been stocked with dog food:

Cherry-Garrard’s summary implicitly presented the missing dog food at One Ton as Scott’s own omission or oversight. However, after Atkinson’s death, Cherry-Garrard discovered further testimonies in 1938 and 1948 corroborating Atkinson’s narrative of Meares’ independent refusal and personal wish to leave Antarctica prematurely (May & Airriess, Reference May and Airriess2014, pp. 269–270; May, Reference May2018a). Given these independent corroborations, Cherry-Garrard now criticised Meares, noting privately in his journal:

By 1948, Cherry-Garrard felt deeply enough about the new discoveries to issue a new “Postscript” to Worst Journey (Cherry-Garrard, Reference Cherry-Garrard and Cherry-Garrard1951). In it, he fills gaps in his earlier narrative: he presents Scott’s orders (from Simpson’s 1912 journal and Evans’s 1921 memoir) and uses them to explain Scott’s decisions concerning dog teams. Cherry-Garrard also exonerates Scott from having understocked his depots and notes his “very strong sense of responsibility” (Cherry-Garrard, Reference Cherry-Garrard and Cherry-Garrard1951, p. 596). However, he refrains from exposing Meares’ refusal to follow orders, preferring instead to place blame chiefly on the expedition’s ignorance of vitamins (pp. 590–1). Although Meares was now deceased, Cherry-Garrard probably wished to protect Meares’ innocent widow.

Unfortunately, Cherry-Garrard did not include these findings within the revised main text of Worst Journey. Doubtless he hoped his “Postscript” would stand as his final word. However, after Cherry-Garrard’s death, someone, possibly his friend George Seaver, chose to omit the “Postscript” from later editions of Worst Journey. Seaver believed that it did “little to add to our knowledge of the circumstances that wrought disaster to a heroic enterprise” (Seaver, Reference Seaver and Cherry-Garrard1965, p. xlii). This omission was a serious misjudgement: Cherry-Garrard’s new evidence and conclusions effectively fell into obscurity. Although Seaver summarised the “Postscript” within his “Foreword” to Worst Journey (pp. xxxix–xlii), most readers either overlooked Seaver’s “Foreword” or dismissed Cherry-Garrard’s revisions based on Seaver’s disparaging verdict. Had Cherry-Garrard’s “Postscript” stood, with its attention to Scott’s 1911 written orders, then both Scott’s thorough professionalism and Meares’ suspicious behaviour would have been generally recognised earlier: the myth that Scott was so unprofessional as to leave “last-minute verbal orders” would not have gained traction in modern histories.

In summary, the cumulative research and evolved understanding resulting in May’s and Lewis’ revised hypothesis in this article can be clearly tracked through May’s post-2012 Polar Record articles (May & Airriess, Reference May and Airriess2014; May & Lewis, Reference May and Lewis2015; May, Reference May2018a). In this new hypothesis, we would exonerate Atkinson of any wrongdoing. We would not blame Atkinson for misrepresenting Scott’s wishes concerning the dog teams, but the man we believe was Atkinson’s source for the false information: Cecil Meares.

Conclusion: our revised hypothesis for events on the Terra Nova expedition (1911–12), and how those events were subsequently interpreted

In conclusion, we present our hypothesis of the expedition’s events and aftermath. If it is adopted by other writers, we would expect our hypothesis to be acknowledged publicly as our specific intellectual property. We would posit this sequence of events:

1. (1) Scott issued written orders, dated October 1911, for dog teams to restock One Ton with dog food (“second journey”) and then “meet” the Polar Party on the Great Ice Barrier during their return from the South Pole (“third journey”). It appears Scott never altered these orders. Although Scott took the dogs further south than originally intended in November–December 1911, Meares himself appears to have recommended or endorsed this decision, given Scott’s 25 November 1911 comment that “Meares says [the meat from] another pony will carry him [his dog-teams] to the Glacier” (Scott, Reference Scott and Jones2006, p. 330). No record exists in Atkinson’s, Simpson’s or Scott’s 1911–12 journals of Scott’s cancelling/curtailing the dog teams’ future movements.

2. (2) Meares returned to base with dog teams in December 1911, having earlier taken too much food from Mount Hooper depot (May & Airriess, Reference May and Airriess2014, p. 265). This created a potential crisis: parties following Meares, man hauling their sledges, would have less food than expected. However, Meares’ not alerting Simpson to this situation is explicable if he, at this point, still intended to fulfil the “second journey” in January 1912 to restock One Ton with dog food for the dog teams to reach Mount Hooper by March 1912. (Meares’ depleting Mount Hooper in December 1911 would therefore be compensated by the dogs’ eventual arrival to meet the Polar Party there in March 1912.) Meares’ behaviour indicates that, in December 1911, Meares still fully intended to follow Scott’s written orders.

3. (3) In January 1912, Meares prepared the dog sledges for a journey to One Ton: his recorded behaviour mirrors Scott’s written requirements for the “second journey”, again indicating that he still followed Scott’s orders. However, on 17 January, the Terra Nova was sighted. It appears Meares now independently misinformed Simpson that Meares’ sledge-journey was unnecessary (“a stock of luxuries”). Meares never restocked One Ton, depriving dog teams of dog food to travel beyond One Ton.

4. (4) We hypothesise that, when Atkinson enquired in 1912 why One Ton lacked dog food, Meares responded that Scott had issued “strict injunctions that the dogs should not be risked in any way” and that “Scott was not dependent on the dogs in any way for his return”. In 1912, it was easier for Atkinson to believe that Scott’s orders had been followed than to make the mental leap of suspecting disinformation from Meares.

5. (5) During the crisis triggered by Evans’ near-fatal scurvy in February 1912, we believe that Atkinson repeated these claims to Cherry-Garrard: they erroneously rationalised One Ton’s lack of dog food as Scott’s order. In 1913, Atkinson repeated the story of Scott’s “injunctions”.

6. (6) We believe that in 1917 Atkinson, upon re-reading Scott’s journals, recognised that Scott’s expectations for dog teams by March 1912 were irreconcilable with the story of Scott’s “injunctions” protecting dogs. Atkinson’s 1919 letter (“[Meares] disobeyed orders”) indicates Atkinson’s belated recognition of Meares’ disobedience.

7. (7) Cherry-Garrard in 1922 repeated that Scott had ordered dogs not to be “risked”. After 1937, Cherry-Garrard heard further testimonies corroborating Atkinson’s accusation of Meares and wrote his 1948 “Postscript” to Worst Journey (unfortunately omitted from post-1952 editions).

8. (8) In Norway, Gran wrote various later-life expedition memoirs. These indicate Gran’s honest misunderstanding of Scott’s 1911 written orders and his belief in Atkinson’s 1912–13 testimony, culminating in Gran’s hypothesis of “Scott’s last-minute verbal orders to Evans”.

9. (9) In the 1970s, the journalist Huntford visited Gran and presumably read Gran’s document “The Race for the South Pole”. In his 1979 double-biography, Huntford repeated Gran’s hypotheses without citing Gran as his specific source.

In conclusion, May reaffirms her original 2012 hypothesis that Scott himself never gave orders in 1911 protecting dogs and that this story was a later misrepresentation of Scott’s actual wishes. What has changed since 2012 is May and Lewis’ present belief that Meares invented this claim rather than Atkinson.

The advantage of this fresh hypothesis (Meares’ false claim to Atkinson in 1912 that Scott had ordered the dogs’ protection) is its conformity to primary evidence documentation, and to Scott’s and Meares’ established patterns of behaviour. Scott left a detailed 1910–12 journal containing self-critical analysis of his own errors, demonstrating personal accountability and transparency. Neither Scott, Atkinson nor Simpson recorded any cancellation/curtailment of the dog teams during 1911. Meares, by contrast, was a professional dissembler from an intelligence background and had earlier avoided accountability for his purchase of poor-quality ponies by giving differing accounts to his colleagues blaming other parties (May & Lewis, Reference May and Lewis2015). Meares also told Simpson an unconvincing story of his supposed plan to waste the dogs’ energies on transporting “a stock of luxuries” to One Ton in January 1912: this appears to be Meares’ falsely rendering trivial what would actually have been a crucial journey, following Scott’s 1911 written orders, to restock One Ton with dog food in January 1912 (May & Airriess, Reference May and Airriess2014). From these men’s previous actions, it is clear that if Scott had cancelled/curtailed dog teams he would have later acknowledged this error in writing, whereas if Meares had not followed Scott’s orders he would have obscured his colleagues’ understanding. That Meares misrepresented Scott’s wishes to Atkinson in 1912 (concealing Meares’ earlier refusal to restock One Ton) is entirely plausible.

The pattern of Meares’ behaviour (specifically, his practised ease in deceiving his colleagues) indicates that his cancelling of the “second journey” in January 1912 was not spontaneous panic (“flying the white feather”, in Atkinson’s over-charitable 1919 assessment), but rather his cool, fully-conscious deception to ensure a premature departure from Antarctica planned since October 1911 (Oates, Reference Oates1911). Meares had previously undertaken active war service and intelligence work. No contemporary evidence exists of Meares’ exhibiting panic or loss of control at base in January 1912; Simpson’s contemporary journal indicates that Meares could convincingly hold his nerve throughout his strange narrative of “Irish Stew, Marmalade and Tinned Fruits”.

That said, we doubt Meares ever intended deliberate harm towards Scott’s Polar Party. We believe that in 1912 Meares, in a blinkered wish to leave Antarctica early, ignored the potential difficulties caused by his taking too much food from Mount Hooper; he probably made the facile assumption that the Polar Party would remain strong and not require the “safety-net” of the dog teams’ progressing further south of One Ton. However, for safety’s sake, Meares could have alerted his colleagues at base to the depleted depot and Scott’s ongoing expectations for the dog teams: the evidence indicates that he did not do so. Meares could verbally misrepresent Scott’s wishes to others in January–February 1912 with impunity, since he planned to leave Antarctica by the time of Scott’s predicted return in March–April 1912.

Had Scott survived in 1912 to learn of all this, Meares would probably have faced punishment. However, the confusion at base and the later news of the Polar Party’s tragic heroism resulted in Meares’ evading scrutiny. Unlike other expedition members, Meares published no account of his actions during 1911–12 and delivered no contemporary journals to the expedition archive: this allowed him the option of later plausible deniability and provided only silence when posterity would wish for his own explanation.

Previous research (May & Airriess, Reference May and Airriess2014; May & Lewis, Reference May and Lewis2015) indicates that Meares hindered the expedition by buying poor-quality ponies, depleting Mount Hooper depot and independently refusing the “second journey”. However, this article’s new hypothesis, that Meares misinformed Atkinson in 1912 that Scott wanted the dogs not to be “risked”, means that Meares would bear greater culpability for the tragedy than previously thought.

We find it inexcusable that Meares did not suffer the small loss of pride necessary to alert Atkinson to the fact that he, Meares, had left inadequate food supplies at Mount Hooper for following parties (especially the final Polar Party) and that he had not replenished One Ton with dog food despite Scott’s expectations for dog teams to progress beyond One Ton to meet the Polar Party. Had Meares explained this discrepancy and raised the alarm, Atkinson would have understood the urgency of meeting the Polar Party further south and sent Wright, not Cherry-Garrard, to One Ton with orders to compensate for missing dog food by using surplus human food or sacrificing weaker dogs. In this hypothesis, dog teams could still have met the Polar Party south of One Ton in March 1912, as Scott had ordered, and rescue could have been feasible.

However, the false claim of Scott’s not wishing to “risk the dogs” recast the missing dog food at One Ton as Scott’s wish. The belief that Scott was “not in any way dependent on the dogs” dulled any sense of danger at base. It reassured Atkinson he could send the inexperienced Cherry-Garrard to One Ton only. It prevented Cherry-Garrard from killing spare dogs at One Ton for meat to enable progress south (Cherry-Garrard, Reference Cherry-Garrard1922, p. 420).

By refusing to stock One Ton with dog food and, we believe, inventing this later claim that Scott did not want dogs “risked” as he was “not dependent” upon them, Meares destroyed any chance of the Polar Party’s being rescued. Those at base in February 1912, no longer understood that the Polar Party still expected support south of One Ton. Tragically, Scott’s journal clearly indicates this expectation: on 10 March 1912 at Mount Hooper, a point 112 km south of One Ton, Scott wrote, “The dogs which would have been our salvation have evidently failed” (May, Reference May2012, p. 88).

We revisit the perennial question: could Captain Scott have been saved? (May, Reference May2012; May & Airriess, Reference May and Airriess2014). The May-Airriess article found fresh evidence that the window of opportunity was narrower than previously thought (due to weather conditions), but that rescue might still have been possible with a practised leader (Wright) who understood that speed was crucial. The May-Airriess article’s new findings were subsequently appropriated (unfortunately uncredited) to attempt to excuse the failure to send dog teams beyond One Ton by claiming that rescue was not feasible. However, if Meares originated the story of “Scott’s strict injunctions” rather than Atkinson, the picture changes. Those “injunctions from Scott” necessarily restricted the scope of rescue efforts at base in February–March 1912. Had all possible variables of weather/animal conditions been perfect for rescue, rescue might still not have occurred: the third journey’s dog drivers might still have remained obediently at unstocked One Ton, unwilling to “risk” the dogs “in any way”, believing that Scott was “not dependent” on them. The potentially harmful impact of this misrepresentation of Scott’s wishes must be explicitly addressed, not concealed.

Our present hypothesis may attract controversy: it is not pleasant to imagine Meares capable of placing his own convenience (in leaving Antarctica prematurely) ahead of his comrades’ safety and survival. However, Meares unquestionably left Antarctica early in 1912: the hypothesis that Meares misrepresented Scott’s wishes to Simpson and Atkinson at base in 1912 to facilitate his early departure is a better deduction from the evidence than a hypothesis that Scott, Atkinson and Simpson all suffered from simultaneous memory loss and supposedly “forgot” to record in their 1911–12 journals a drastic alteration of expectations for the dog teams.

The study of history places priority on primary evidence over secondary. We believe that the primary evidence indicates that Scott never gave “injunctions” protecting the dogs, that Simpson, Atkinson and others were deceived by Meares during 1910–12, and that Meares independently disregarded the Polar Party’s needs. Corroborative contemporary journal evidence indicates that Scott himself never changed his 1911 written expectations for the dogs, so this 1912 anecdote of Scott’s “strict injunctions” protecting dogs was an invention at base. Its inventor had to have been either Atkinson or Meares; given the multiple corroborated accusations against him, Meares is the most credible culprit.

If our hypothesis is challenged, we would expect challenges to address the core issues directly, such as the absence of reference to Scott’s cancelling/curtailing the dog teams’ movements for 1912 within Scott’s, Atkinson’s or Simpson’s 1911–12 journals. (That all three journals independently corroborate the absence of any such order is itself evidence that Scott never gave any such order.) Challenges should directly address Meares’ leaving reduced food at Mount Hooper depot and not alerting Simpson to this potential crisis. Challenges should fully quote and directly address Meares’ dishonest-sounding announcement, recorded in Simpson’s handwritten 1911–12 journal, that Meares supposedly planned to transport “luxuries like Irish Stew, Marmalade and Tinned Fruits” 490 km, to One Ton and back, in January 1912. (Simpson’s 1910–12 handwritten journals, preserved for posterity and delivered to the expedition archives, indicate Simpson’s personal transparency and accountability (May, Reference May2018a).) Challenges should address the fact that Atkinson, who related Scott’s “strict injunctions” protecting dogs in 1912–13, recorded no such “injunctions” from Scott in his 1911 journal. Challenges should address the fact that Atkinson and Cherry-Garrard, who initially reported the story of Scott’s “strict injunctions”, later believed (with two others) that Meares had “disobeyed [Scott’s] orders” (May & Airriess, Reference May and Airriess2014). Challenges arguing Meares’ honesty and loyalty should directly address his habit of false narratives, his previous failures to follow Scott’s orders (buying poor-quality ponies in Siberia, refusing to descend into a crevasse to retrieve dogs on 21 February 1911), and his premature departure from the expedition. This list of crucial issues should not be ignored in any future challenge.

The question of primary evidence is crucial, since a recent article employs Gran’s unsubstantiated later-life hypothesis to accuse Evans of failing to deliver Scott’s supposedly re-instated original orders in 1912 (Turney, Reference Turney2017). However, Scott’s, Simpson’s and Atkinson’s 1911–12 journals indicate no cancellation of Scott’s 1911 written orders. This 2017 allegation of Evans’ supposedly failing to deliver Scott’s re-instatement remains unfounded: it disregards both the 2012 criticism of this Evans-related anecdote (May, Reference May2012) and the primary evidence implicating Meares first presented in 2014 (May, Reference May2018a; May & Airriess, Reference May and Airriess2014).

We believe current confusion in the expedition narrative originated from expedition members, then historians, taking accounts on trust rather than tracing claims to their earliest sources. Modern theories of Scott’s sentimentality towards animals during the depot journey, and his last-minute “verbal orders” to Evans, apparently originated from Gran’s honest misunderstanding of Scott’s intentions and writings. Gran sincerely intended to write secondary history: aware of the problems within the conventional expedition narrative, he attempted to “fill in the cracks” with hypothesis as secondary historians would. Unfortunately, it seems Gran supplied his own explanations without understanding the dangers of unsignalled hypothesis. Subsequently, when Huntford introduced Gran’s hypotheses into English-language polar scholarship, they were not checked against primary evidence: Gran was not explicitly credited as source. Consequently, Gran’s hypotheses hardened into unquestioned “fact”.

While the causes of the Terra Nova expedition disaster are multi-factorial (involving weather conditions, and the strength of animals and men), scholars have believed for decades that vital factors included Scott’s issuing unclear orders, placing priority on animals over humans and giving “strict injunctions” to Atkinson protecting dogs. This narrative damaged Scott’s posthumous reputation. Primary evidence now indicates that Scott valued human life over animal, that in October 1911 Scott issued written orders which he expected to be followed, that Atkinson never received “injunctions” from Scott protecting dogs, and that during their return on the Barrier, Scott’s Polar Party were denied the possibility of assistance from dog teams (upon which they depended for survival, considering Oates’ incapacitation) because understanding of Scott’s wishes had been corrupted at base.

The evidence indicates that the source of this corruption was Cecil Meares. We now believe that, in January 1912, Meares opportunistically exploited his colleagues’ trust to avoid a vital errand and then subsequently disseminated disinformation to Atkinson that Scott no longer required dog teams to meet him south of One Ton. Had the testimonies from Cherry-Garrard, Simpson and others been re-examined in the 1970s, the evidence pointing to Meares’ disobedience and disinformation would have probably been discerned four decades ago, and Gran’s later-life hypotheses recognised as sincere but mistaken. Unfortunately, in Huntford’s (Reference Huntford1979) narrative, all blame was attributed to Scott, Meares was presented as blameless and Tryggve Gran’s later-life hypotheses were repeated as cast iron fact.

In 2015, May wrote:

Huntford’s (Reference Huntford1979) repetition of Gran’s post-1945 hypotheses without citing Gran as his specific source illustrates why historians’ narratives must credit others’ hypotheses. Historians need to cite sources clearly: by doing so, they allow future scholars to make a distinction between primary evidence and later secondary hypothesis.