Introduction
   The following notes are intended not as a rigid set of rules but as a guide for future Antarctic travellers: common sense must be used in their application. Systematic breeding, training and driving of dogs are essential in an organization as large as the Falkland Islands Dependencies Survey, but modifications must naturally be made to suit particular requirements.

   Among the first members of the Survey, only two - Captain A. Taylor, R.C.E. (Port Lockroy, 1944-45 and Hope Bay, 1945-46) and Surgeon-Commander E.W. Bingham, R.N. (Stonington Island, 1946-47) - had had previous sledging experience. With the exception of a few modifications derived from Taylor's experience, Bingham's technique of dog driving was followed in later years (Bingham, 1941).

    All the sledge dogs originally used were brought from Labrador in 1944 and 1945 (Bingham, 1947a, p. 24, 31; James, 1947, p. 40). Few of these still remained in service in 1948 and 1949, but their progeny, born and bred in the Antarctic, proved larger in size and better-tempered than dogs brought from Labrador (James, 1947, p. 42).

Breeding
   To obtain, and to retain in successive generations, a high standard of efficiency as a traction animal, sledge dogs must be bred with care. With very few exceptions domestic animals have been improved immensely by careful breeding through the centuries. Animals possessing special qualities, such as size, strength, milk yield and colour, have been constantly selected and mated with discrimination. This preserves and intensifies the genetic factors which, in good environmental conditions, produce in the mature animal whatever is specially valued. Although sledge dogs have been used domestically for many generations, deliberate genetic research has played little part in breeding. Further improvement is probably attainable in physical conformation, stamina, and physiological efficiency. In the Falkland Islands Dependencies Survey, where the whole dog population consists of not more than about 300 descendants of the few dozen animals originally imported from Labrador, conscious care in breeding is essential if efficiency is to be maintained. Continued breeding within this small population means that all the dogs must inevitably, after a few generations, be related to one another, leading to a concentration of genetic qualities, good or bad. So-called in-breeding, its popular meaning being the close and haphazard breeding within a small population, commonly leads to general deterioration. On the other hand, a very considerable degree of breeding between close relatives under proper control is the standard method of "fixing" in the breeding qualities of high value. Careful breeding is thus essential if the quality of the Falkland Islands Dependencies Survey stock of dogs is to be maintained, and the introduction of new blood may be necessary in the future. New blood, however, will not ensure high quality unless there is conscious discrimination.

   The following traits are desirable in dogs and bitches selected for breeding: good physique and stamina; ability to lead a team; intelligence; good sense of direction; readiness to pull, even under adverse conditions; big feet and long, well-built legs; broad chest and alert stance; broad head and muzzle. Bitches selected should be good mothers and capable of suckling their pups without aid. Some dogs appear to attain a good physical condition while at the base but rapidly lose condition while sledging. It is therefore inadvisable to breed from them. It is better to breed dogs with short coats. At Hope Bay a particularly "shaggy" strain resulted from repeated and indiscriminate in-breeding of dogs with thick woolly coats, the length of the underwool or fur fibres being almost the same as that of the coarser guard hairs. Drift snow tends to adhere to a "shaggy" coat, so that a dog becomes encased in ice, or has large balls of ice hanging from its coat. At Hope Bay a dog had to be destroyed because the excessive weight of ice caused the skin on its back to tear, exposing the flesh. Several of the "shaggy" dogs at Hope Bay were subsequently sheared like sheep, but they still became "iced up".

   To ensure that a bitch is served by a particular dog, both should be chained in close proximity during the heat period of the bitch. A bitch on heat in the field can cause chaos, and at night should be tethered separately some distance from the rest of the team. At times it is desirable that bitches should not be served, for instance during the first heat. For this purpose Bob Martin's "Antimate" has been used with some success.

   Although sixty-eight^l pups were born at Stonington Island in 1947-48, only three were kept because rearing entails much extra work, especially when the majority of the staff are absent from the base. For this reason it is considered preferable to breed pups at small stations, where more time is available and food is more plentiful. On the other hand, dogs bred at a sledging base can be trained at an early age to the ways of those who are to drive them.

   All pups born in the field were destroyed soon after birth. Whenever pups are not required they should be killed at the earliest opportunity, preferably within a few hours of birth. If bitches are allowed to suckle even a single pup for several days before it is destroyed, they invariably suffer from enlarged and inflamed mammary glands. This condition may cause abscesses which are difficult to cure, and keep a bitch out of her sledging team for an unnecessarily long period; occasionally it may even be necessary to destroy her. If the pups are taken away at birth, lactation ceases within two or three days, any milk being resorbed. In some cases bitches whose pups have been destroyed at birth retain condition and may be worked after a few days' rest. Others lose condition irrespective of whether they suckle a little or not.

Weaning
   In 1948-49 thirteen pups were successfully reared at Stonington Island. The bitches were allowed to suckle them until the beginning of the third week. During this period the bitches' normal ration was supplemented with a special puppy mea1.² Weaning then began, first by introducing the pups to Nestle's sweetened condensed milk mixed with an equal quantity of water. Within a week they drank from a saucer, but only if they were not allowed to feed from the mother. At this stage they were fed four times daily. Before the pups were fed in the morning the bitch was chained apart to prevent her suckling them during the day. After the last feed of the day the bitch was put back with the pups to keep them warm during the night. In the fourth and fifth weeks pups were fed twice daily on a thin porridge-like mixture of puppy meal enriched by condensed milk. The condensed milk was later omitted, and the food made a little thicker. The bitch was taken away from her pups after six to seven weeks. Feeding on meal continued until such time as the pups could eat minced seal meat and liver without difficulty.

   Should a bitch be unable to suckle pups, they may be reared from birth on Nestle's sweetened condensed milk, mixed with water in the proportion of two parts to one. For this purpose a baby's bottle and rubber teat may be used.

   In the first few weeks of weaning it is preferable to feed pups a little thrice daily rather than to overfeed them less frequently. Overfeeding may cause intestinal trouble (intussusception), which resulted in the death of three pups at Stonington Island. One meal a day is possible after eight to ten weeks. At this time it is desirable to provide water for the pups until they learn to eat snow.

Kennels and pens
   Drift-proof canvas-covered kennels were provided for bitches with litters and for young pups being weaned. At night, and during blizzards, the entrances were covered with sacking to prevent drifting snow from filling the kennels and turning to ice on the floor.

   Pup pens were made of netting-covered tubular frames, 4 ft. 6 in. x 6 ft. Until the pups were three months old the pen was 4 ft. 6 in. high, but when they became more adventurous and began to climb out, the sections were turned on end to form a pen 6 ft. high. A kennel was built into one side of the pen.

Tethering
   At the beginning of 1947 the majority of dogs at Hope Bay were roaming freely. During several months of freedom they had formed into groups, each with its own "king" dog. There was, however, one "super-king" dog acknowledged by minor "kings". There were endless fights, and it soon became necessary for a member of the party to be constantly on watch to prevent injuries. Eventually the dogs became unmanageable, especially when sledging began, and they had to be chained. Another reason for doing this was the danger that they would stray (James, 1947, p. 41).

   A.R. Glen (1939, p. 185) prefers the use of pens to that of spans and chains. If there is relatively little snowfall or drift, and few dogs, this is a practical proposition, but at a base like Hope Bay where there were 134 dogs and pups in 1947, it was impractical. The same applies to Bingham's use of deadmen for tethering (Bingham, 1947b, p. 44).

   Tethering lines of steel wire cable (often called spans) to which 6 ft. dog chains were attached with bulldog grips at intervals of 15 to 18 ft., were used at Hope Bay and Stonington Island in 1948 and 1949. All chains could be fitted with two swivels and a swivel clip-hook, to prevent tangling. Each cable, accommodating nine or ten dogs (usually a team), was 3/4 in. in diameter and each end was firmly picquetted. The cables were carefully inspected from time to time so that fraying could be prevented, or noticed before the cable snapped. Occasionally the nuts on a bulldog grip worked loose and the friction of a grip sliding along the cable led to rapid fraying. Pairs of dogs often came too close if the bulldog grip slipped, and fights resulted. This trouble was remedied by unlaying two strands of the cable and inserting the bulldog grip through the gap.

   Before 1948, dogs at Stonington Island were tethered to deadmen buried in the snow at intervals of 15 to 20 ft. Deadmen were also used when a few dogs had to be tethered on sea ice during winter. Disadvantages are that a dog can easily pull them up in summer, and in winter they are difficult to dig out.

   Light spans and chains have been suggested for tethering at night while sledging (Bird & Bird, 1939, p. 182). They were used by Hope Bay parties in 1946, but since 1947 the method described below has been used. Spans are cumbersome, and are unnecessary when dogs are well trained.

   On the whole, the method of tethering adopted at Hope Bay and Stonington Island in 1948 and 1949 proved satisfactory. The comments of Bingham (1947b, p. 44) on the tethering methods used by James (1947, p. 43) were fully supported by later experience. A centre trace formation was generally used, the dogs were unharnessed and left tethered to the traces by their collars. Some dogs were picquetted separately for a specific reason. When fan trace formation was used the dogs were tethered in threes by tying three adjacent traces together with a slip knot, and attaching them to a picquet.

   The use of ice-axes for picquetting (James, 1947, p. 43) is wrong, and resulted in the breakage of twelve hafts in two seasons at one base alone. Steel angle-iron or spiral picquets are much more efficient and will hold a team without difficulty. In soft snow a deadman, made from a ration-box lid or a picquet laid horizontally, should be used. A pair of crampons, buried points downward, are equally satisfactory.

   If dogs are to be tethered, it is essential to provide a suitable collar for attaching the clip-hooks to the chains. Clip-hooks must be snapped on to the collar itself and never on to the D-ring of the collar. Constant pulling on the D-ring very soon ruins a collar, especially when dogs are tethered to cables at base for long periods. Collars may be of spliced cod-line or marline, flat leather, or a round rope-core covered with leather. Each type has a particular use. Unlike most cheap leather collars, neither cod-line or marline freeze in winter. Strap leather collars can easily be made if buckles and leather are available: they must not be too wide, or they will catch in the clip-hooks and wear quickly in one place. Strap leather collars were found most successful, because they slide easily round the neck, and neither tangle in the ruff nor chafe the skin under the throat. Round cross-section leather collars are preferable but they are too expensive. The kind of collar used is largely a matter of individual preference, but the use of choke collars is inhumane.
 
   Whenever possible, dogs should be tethered on snow, because this keeps them cleaner.

Harnesses
   Harnesses used by the Falkland Islands Dependencies Survey are made up of 11 in. (flat measurement) tubular lampwick, usually hand-sewn with spun yarn. Single thickness 11 in. lampwick should never be used, as it stretches, cuts into the dog's coat, and chafes the skin. Well-fitting harnesses are essential for efficient hauling, and for dogs of normal size (of about 85 to 90 lb.) the measurements are shown in Fig. 2. If necessary the measurements of A'B' and A'D' may be increased or decreased by 1 in. or, very occasionally, more. If a dog can easily slip out, the harness is at fault, generally because it fits badly under the forelegs. Several modifications to the original pattern (Fig. 1) may be made: for instance, the back straps (A'E') of the harness can be nipped at C'; the cross-piece can be lengthened by increasing it from 6 to 8 in.: the shoulder straps (B'D') can be increased to 12 in. instead of the customary 10 to 11 in., and A'B' can be reduced from 13 to 7 in. The reduction in length of the shoulder strap and the corresponding increase in length of the cross-piece not only provides a better neck aperture but a more comfortable fit on the shoulders. This point is illustrated in Fig. 3, where the original and modified patterns of neck aperture are shown. The cross formed by the two shoulder straps should rest on the centre of the chest, as shown in Fig. 4.

   Some dogs delight in extricating themselves from harness, however well-fitting it may be. To prevent this a small loop-and-toggle belly-band may be fitted at F' (Fig. 2).

   If the dog's name is embroidered or inked on the cross-piece, identification is simple, and anyone who knows the dogs by name can harness the team. Coloured tags, until 1948 at Stonington Island, were of little value for this purpose because invariably the team driver was the only person who knew to which animal each particular harness belonged. Lampwick harnesses may freeze: the cross-threads of some types of lampwick crack and fray, leaving a tangled mass of longitudinal threads. This can be prevented by soaking in paraffin prior to use, a treatment which does not harm the dogs' skin.

   As Glen (1939, p. 186) comments, dogs seldom chew harnesses and traces when they are well-fed and comfortable. As soon as they become very hungry, or entangled in the traces, they revert to chewing.

Traces
   Although the method of driving and consequently the type of trace used should depend on surface conditions and the nature of the terrain, a driver usually prefers one particlar method and type of trace, which he must however be prepared to change with changing conditons. Of the two methods used at Stonington Island in the sledging seasons from 1948 to 1950 - centre trace and paired fan³ - the former was used by every driver with great success, and was preferred owing to its simplicity.

   Centre trace. The centre trace was standard and accommodated nine dogs. Extensions to take two additional dogs could easily be added. The trace, made of 11 in. sisal rope⁴, comprised three sections 7ft. 6 in. long and one 9 ft. long, with 1 in. metal rings spliced between each section. Although galvanized steel rings were generally used, stout brass rings are preferable because they do not rust and wear the splices. The 9 ft. section had a 3 to 4 in. loop spliced in the end attached to the main trace. Reinforcement of this loop with a canvas sleeve prevented wear. 

   Side traces. The best length for side traces is between 2 ft. 3 in. and 2 ft. 6 in. Two types were used, both made of tarred marline.
   The first type (Fig. 7) consisted of a loop spliced at each end of a suitable length of line, with a catspaw knot attaching the clip-hook to the side trace, and the side trace to the ring of the centre trace. The loop of the catspaw often slips over the clip-hook swivel, preventing it from working correctly, and may cause the trace to unlay.

   The second type (Fig. 6) is made of splicing a loop at one end of the line and a clip-hook at the other, care being taken not to make the clip-hook splice too small. The side trace is then attached to the centre trace by a catspaw knot.

   If the loop at the clip-hook end of the side trace wears through, the trace can be shortened and the clip-hook respliced. The length of the leading dog's trace is determined by the team driver according to its habits.
   On all sledge journeys spare traces, complete with clip-hooks, were carried to avoid delays when traces broke.

   Main trace. With the exception mentioned below, each sledge was equipped with an ordinary one-piece main trace, to which the centre trace was attached by two adjacent karabiners. The karabiners, however, were found to wear both the main and the centre traces at attachment points. To prevent this a canvas sleeve was sewn over the main trace. Rubber hose has also been used for this purpose.

   A modified type of main trace, made in two parts (Fig. 8), proved easier to operate, even when iced-up. A loop was spliced at one end, and when the centre trace loop was slipped over the other end, a double sheet bend was tied; reknotting in a different place each time distributes wear evenly over attachment points.

   A square toggle and loop, similar to that used by Bingham (1941, p. 79), can be used, but does not prevent continual wear at attachments.

   Fan traces. Sets of paired fan traces for seven, nine and eleven dogs were available at Stonington Island, and were used for the greater part of the 1947-48 season, but hardly at all in 1948-49 and 1949-50. The most suitable measurements for individual traces were two of 9 ft., two of 15 ft., two of 21 ft., two of 27 ft., and one of 33 ft.

   The difference in length between pairs of traces is 6 ft., and a square-toggled extension trace is used for attaching them to the main trace. This type of extension (Bingham, 1941, p. 379) to the main trace is important, because it allows easy unravelling of the fan throughout the day without the possibility of dogs escaping.

Feeding
   During the summer it was customary at base to give each dog approximately 4 lb. of seal meat without blubber every alternate day. Blubber was not given to dogs during the summer because it always passes straight through them, and invariably finds its way on to their coats, which become badly matted.

   During the winter the dogs at base usually received every alternate day approximately 6 lb. of meat and blubber, of which a third was blubber. In 1948 very few seals were killed at Stonington Island after the beginning of March.⁵ The resulting shortage of meat necessitated the introduction of different forms of feeding. Blubber which had been saved during the late summer was cut into squares weighing approximately 1 to 1 1/2 lb. One of these, together with 2 lb. of Bovril dog pemmican, constituted one meal. Because dogs appear to need oil or blubber, especially in winter, stock fish (dried cod) was used together with blubber as a meal on alternate days. Owing to the limited quantities of seal blubber and stock fish available at Stonington Island in 1948, pemmican alone was fed to the dogs for long periods. In order to provide some variety, it was finally decided to feed pemmican daily for three days followed by stock fish and blubber on the fourth day to last two days. Occasionally stock fish were fed together with 1 lb. of pemmican on alternate days.

   On all sledge journeys the dogs were given 1 lb. of pemmican a day. Double feeds were given whenever signs of fatigue appeared. At Stonington Island in 1948 it became the practice to reserve a small stock of seal meat, as a special contribution to the dogs' diet in the ten days before a winter sledging journey began; this provided four good meals, which perhaps helped the dogs to withstand the rigours of winter sledging. Lack of stamina during any particular winter journey may be accounted for by poor feeding prior to the start. While the dogs were away on such a journey, a number of seals were often killed in Neny Fjord, and on their return to base the dogs were given seal meat and blubber for several weeks until the next journey began. From the beginning of October until the beginning of March the following year they were fed on seal meat. This, together with constant exercise, made a great improvement in their general condition.

   The failure to revictual Stonington Island in the summer of 1948-49 caused a shortage of dog pemmican, and in 1949 precautions were taken to ensure that the dogs would have an adequate supply of seal blubber throughout the winter. Again seal meat and blubber were stored for feeding the dogs before the winter journeys. No signs of any dietary deficiencies were observed in the ensuing winter.

   As prolonged feeding on pemmican in the field may be detrimental to the dogs' health, it is desirable to give them seal meat once or twice every ten days, especially if heavy loads are being hauled and the duration of the journey exceeds thirty days. While sledging during winter, dogs were given more than 1 lb. of pemmican daily whenever sufficient was available. An extra 1/2 lb. of pemmican on alternate days is preferable, but it is usually difficult to carry the additional load. K.S.P. Butler did this on the main southern journey in 1947-48 down the west coast of the Weddell Sea, and all his dogs returned to base in excellent physical condition after 105 days in the field.⁶ Butler was able to give his dogs 1 1/4 lb. of pemmican daily because his party had air support.

   James (1947, p. 41) and Bingham (1941, p. 373) both advocate hot meals for dogs at base in winter. Once again, this depends upon the number of dogs; with large numbers it is out of the question, but for sick dogs, or bitches with litters, it is certainly advisable. In summer, when snow is not available at the tethering place, dogs must be given water for drinking every day.

   At Stonington Island in 1948 a total of 3260 lb. of pemmican was used for feeding at base in the winter, and approximately 5000 lb. during sledge journeys. In addition, fifteen bales of stock fish and some eighty seals were used at base. Similar quantities were consumed in 1949. 

Training
   Pups were accustomed to harness by walking them round, thus allowing them to exercise their natural instinct to pull. The next step was to run them in a quiet team, at first alongside a bitch. Normally a pup will pull with vigour, and within half an hour will become completely exhausted. Such training for ten successive days should enable a pup to take part in journeys. Stamina for pulling heavy loads over great distances for long periods develops only after a pup has been in the field for several months.

   Pups can be trained to pull in any of the standard formations, but it was found most successful to use a centre trace because when several pups are pulling in paired fan formation chaos usually results. Once discipline has been instilled, training can continue using paired fan traces. With several older dogs in the leading positions, usually two young dogs are inspanned in the centre of the team (Fig. 9). An additional neck line from the pup's collar or shoulder straps to the centre trace ring ahead generally prevents the pup from dawdling, pulling out of harness backwards, or otherwise causing trouble. A belly-band (Fig. 2) may be used in addition to the neck line to prevent the pup from slipping out of the harness.

   In training leaders, care should be taken to accustom the dog to a forward position in the team at an early stage; later the dog should be put up alongside the leader until such time as he will answer all commands and is proficient in leading. Then he may lead his own team. 

The words of command used by the Falkland Islands Dependencies Survey are corruptions from arctic terms. In order to standardize them the following terms were used at Stonington Island:

To start "UP DOGS, WEET"
To stop "A-a-a-a-"
Turn right "AUK, AUK" ("au" pronounced as "ou" in loud)
Turn left "I-r-r-re" (with a long rolling "r")

   With well-trained teams it is unnecessary to repeat commands, unless turning right or left, when the command must be repeated until the dogs are heading in the required direction.

Acknowledgements
The writer gratefully acknowledges information about dog-breeding contributed by G.C.L. Bertram, and constructive criticism by B.B Roberts and V.E. Fuchs.

References
BINGHAM, E.W. (1941). Sledging and sledge dogs. Polar Record, Vol. 3, No. 21, p. 367-85.
BINGHAM, E.W. (1947a). The Falkland Islands Dependencies Survey, 1946-47. Polar Record, Vol. 5, Nos. 33/34, p. 27-39.
BINGHAM, E.W. (1947b). Comments on D. James's article on the sledge dogs of the Falkland Islands Dependencies Survey, see below. Polar Record, Vol. 5, Nos. 33/34, p. 43-44.
BIRD, C.G. & Bird, E.G. (1939). The management of sledge dogs. Polar Record, Vol. 3, No. 18, p. 180-84.
GLEN, A.R. (1939). Comments on Messrs Bird's article on the management of sledge dogs, see above. Polar Record, Vol. 3, No. 18, p. 184-87. 
JAMES, DAVID (1947). The sledge dogs of the Falkland Islands Dependencies Survey, 1945-46. Polar Record, Vol. 5, Nos. 33/34, p. 40-43. 

Postscript
   The first draft of the article reprinted above (by courtesy of the Editor of The Polar Record) was written at Stonington Island in 1949 as a "base report" on the state of the dog population. In due time, and after countless redrafts, it was published in 1952. Many articles on huskies and dog handling, both Arctic and Antarctic, had previously appeared in The Polar Record, and the idea behind the publication of this article was to reflect the dog-handling techniques current in the Antarctic in 1950. David James had already written about the dogs used by Operation Tabarin, and Surgeon Commander E.W. Bingham had recorded aspects of dog-handling in the first days of the Falkland Islands Dependencies Survey.

   In the early years of FIDS there was total reliance on dogs for transport. All of the long reconnaissance journeys were done by dog sledge with virtually no support, except for a few field depots. At one time, to say you had been to the Antarctic was tantamount to saying you were an expert dog driver - far from the truth, because there is much more to successful dog-driving than meets the inexperienced eye!

   However, field-work techniques were changing. The first aircraft arrived in the Antarctic to support FIDS field work and it was not uncommon to see a two-man sledging unit, complete with dogs, being loaded aboard one of these aircraft. Then came the first motor toboggans which were initially used on an experimental basis until their reliability was thoroughly proven. Once field work had been fully mechanized, the dogs were slowly relegated to a back-up role and a small breeding population of about 40 dogs was maintained at Adelaide Island. This decision precipitated countless fierce arguments between the "pro-dog" and the "pro-vehicle" camps, and to this day, even in England, these same arguments go on between ex-Fids. This is still one of the main talking points at BAS Club Reunions!

   Times have changed; all of us must move with advancing technology, but for how much longer will the Survey's dog population survive. It is difficult to appreciate fully that 26 years have passed since this article was first drafted.
 

End Notes
¹ Out of sixty of these pups (in eight cases sex was not identified before destruction) forty-five were males and fifteen females.
² "Casco Puppy Meal", a proprietary dog food, was available at Stonington Island.
³ "Paired fan" (trace lengths in pairs except for single leader) is a new term suggested for what has previously been known by a variety of terms such as "modified fan", "modified fan hitch", "British Graham Land Expedition method" or "Mackenzie River method".
⁴ Sisal rope was used because neither manilla nor hemp was available. Manilla and hemp wear better and last much longer than sisal.
⁵ No seals were brought from the Argentine Islands in February 1948, because the John Biscoe was unable to make a second run round the Falkland Islands Dependencies Survey bases.
⁶ This period includes the ninety-nine days of the main journey and six days travel immediately preceding. Seal meat was fed at intervals throughout this period.