NATURE 



[June 2, 1904 



own, and one of his (i.e. Gould's) early associates, that I 

 first, about 20 years ago, became aware of his discovery of 

 many evidences of glaciation in Tasmania." 



I have recently found a Parliamentary Paper, issued in 

 i860, in which Gould describes his recognition of Glacial 

 action in some of the high valleys of central Tasmania. 

 The passage is as follows (" A Report of the Exploration 

 of the Western Country by Mr. Gould," Pari. Pap., 

 Tasmania, i860, No. 6) : — 



" In the Cuvier Valley I was struck, both in going and 

 in returning, by the similarity to the terminal moraine of 

 a glacier presented by an enormous accumulation of boulders 

 which chokes the lower end of the valley, and, somewhat 

 like a dam, extends completely across it, with the exception 

 of the point where it is broken through by the river." 



I am glad, therefore, to be able to give to Gould the 

 credit of having published the discovery, which in my paper 

 I could only quote as a verbal tradition. 



The Cuvier Valley is one day's journey west of Lake St. 

 Clair. -A. hut, five miles due west of the top of Mount 

 Arrowsmith, occurs in it. J. W. Gregory. 



The University, Melbourne, Victoria, April 25. 



The Origin of the Horse. 



In your issue of May 19 (p. 53) Prof. T. D. A. Cockerell 

 refers to Eqiitis caballns celticus, Ewart, as " still surviving 

 in the pure state in Iceland." Prof. Ewart, in his paper 

 on " Tile Multiple Origin of Horses and Ponies," says that 

 ■' the few pure specimens of the Celtic pony survive " in the 

 north of Iceland. I tal-:e it that Prof. Ewart does not mean 

 that the northern Icelandic breed of ponies is a pure one, 

 but only that certain individuals of this breed exhibit the 

 " Celtic " characters in a very marked degree. In a recent 

 paper (Proc. Camb. Phil. Soc, vol. xii., part iv.) Mr. 

 F. H. .\. Marshall and I have brought forward both 

 historical and zoological evidence for the mixed origin of 

 the Icelandic pony. It is perhaps worth noting that the 

 people of north Iceland still claim a social superiority over 

 those of the south as being descended chiefly from the 

 second body of colonists which reached the island. In 

 considering the origin of different breeds of the domestic 

 animals ethnological considerations are often important, 

 and, conversely, the examination of local breeds may some- 

 times throw light on ethnological problems. For example, 

 in the Malay Peninsula the breed of dogs owned by the 

 majority of the jungle tribes usually classed as Sakais 

 differs from that of the Malay pariah, which has recently 

 been adopted in some cases by Semang tribes and also by 

 those Sakais who live in close intercourse with the Malays. 

 The pariah seems likely to oust the Sakai dog completely, 

 and I am not aware that any zoologist has yet made a 

 detailed examination of the latter, which shows certain 

 resemblances to the local race of Cyan ruiilans. 



Of course, investigations into the ethnological distribu- 

 tion of animals must be made with the very greatest care, 

 for not only may one breed oust or swamp another, but 

 the characters of a single individual may prove so dominant 

 that they may prevail in a great number of cross-;bred 

 descendants, and so change the character of a breed in a 

 very short time. This has recently happened in the Fjieroe 

 Isles. As we know from the statements of Landt 

 (" Description of the Faroe Islands," 1798), there were at 

 least two distinct breeds of dogs in these islands at the 

 end of the eighteenth century, one resembling the modern 

 Danish hound, but smaller ; the other a short-legged, rough- 

 haired terrier. The two breeds can still be traced on some 

 of the islands, notably on Naalsoe ; but in the neighbour- 

 hood of Thorshavn, the capital, great alteration has taken 

 place quite recently. Some ten or twelve years ago a 

 Danish governor introduced a well-bred dachshund dog, 

 which inter-bred with the native bitches. In 1903 I could 

 hardly find a single dog in the town which did not show 

 traces of dachshund ancestry — short, bent legs, long body, 

 &c. — more or less marked. The in-bred highly specialised 

 individual has proved prepotent when crossed with the more 

 or less generalised types which, judging from the statements 



NO. 1805, VOL. 70] 



of Lucas Debes (1623-1670) and Landt, have been some- 

 what cross-bred for at least two and a half centuries. A\'e 

 are apt to forget factors of the kind when discussing the 

 breeds of domestic animals, and also when investigating 

 the different races of men, but it should be remembered 

 that they are of the very greatest importance in both lines 

 of inquiry. Nelson .\nnand.^le. 



34 Charlotte .Square, Edinburgh. 



Insular Races of Animals and Plants. 



If we accept the view that species are such by virtue of 

 segregation, and consider subspecies to be groups as yet 

 imperfectly segregated, we seem logically bound to regard 

 insular forms as valid species. According to this way of 

 loolcing at the matter, a subspecies is in biology what a 

 peninsula is in geography, while a species corresponds to 

 an island. Hence it follows that many subspecies are far 

 more widely distributed and for most purposes more im- 

 portant than many distinct species ; just as many peninsulas 

 are more important than the small islands off their coasts. 



While it appears illogical to treat insular races as sub- 

 species, there are difficulties in the way of regarding them 

 all as distinct species. In former years, the most distinct 

 were so recognised, and the others were simply ignored. 

 This practice, while it smoothed the way for the systematist, 

 deprived us of the use of a large body of facts of the 

 greatest possible interest to the evolutionist, and the time 

 has come when it must be given up. As a result of the 

 new methods, the number of " species " recognised is in- 

 creasing very rapidly, as shown, for example, by the de- 

 scription of seventy new Malayan mammals in a single 

 paper by Mr. G. S. Miller, jun. Many of the " species " 

 described in this paper are excessively similar and yet 

 distinguishable, and inhabit different islands. It is evident 

 that one could take a map of the Malay Archipelago and 

 prophesy with some degree of accuracy the number of 

 insular species of Mus and some other genera awaiting 

 discovery by simply counting the islands, eliminating those 

 too closely adjacent. In mountain regions something of 

 the same sort is found, the tops of the mountains or moun- 

 tain ranges serving the same purpose as islands. For fresh- 

 water organisms, lakes and river systems afford similar 

 phenomena, as shown, for example, by the races or species 

 of Salmonida?. 



The objections to the recognition of all these isolated 

 forms as valid species are two. First, their extreme 

 similarity in many instances, and second, the specific name 

 does not indicate the immediate relationships of the form. 

 It has seemed to me that these difficulties might be over- 

 come by the recognition of a new category, for which the 

 name " idiomorph " suggested itself. This name may be 

 objectionable on account of the term idiomorphic, used in 

 crystallography, and it is probable that someone can think 

 of a better. If it is accepted, it may be abbreviated to 

 " id.," as " var." is written for variety, and " subsp." for 

 subspecies. 



To illustrate the different methods, we may take certain 

 bats of the genus Chilonycteris, found in the Greater 

 Antilles, using the facts recently published by Mr. Rehn. 



C. macleayii group, 

 i. (Species). ii. (Sub-species.) iii. (Idiomorphs). 



C. grisea, Gosse. C. macleayii grisea. C. {.macleayii id.) 



macleayii. C. tnnclcayii. 



macleayii /nil- C. {macleayii id.) 

 ginosa. fuligifiosa. 



macleayii in- C. {macleayii id.) 

 /lata. iitflala. 



Cuba ... C. macleayii. Gray. C. 

 Haiti ... C./K//f'»M<i, Gray. C. 



Porto Rico C. itijiala, Rehn. C. 



C. parncllii group. 



... C. pamellii.QrSiy. C.pamellii. C. J>arttellii. 



Cuba ... C ^o£7//[/, Gundlach. C. pantelUi boothi. C. (/ar/tellii id. 



Haiti ... C. fl C.Mrnellii' C. (.parncllii M.)'. 



Porto Rico C.porforicensis. C. parnellii portori- C. (parnellii id.) 



Miller. censis. porioricensis. 



I Doubtless exists, but not yet dis 



ered. 



