620 



SCIENCE 



[N. S. Vol. XXVII. No. 



isolated as ever, and its distribution is still 

 disconnected, indicating great antiquity and 

 partial extinction. The PeripatidsB include 

 two genera, Peripatus and Eoperipatus. The 

 former is neotropical, with the exception of a' 

 single species (P. tholloni) found in the 

 French Congo. The latter consists of three 

 species, two from Malacca, the other Su- 

 matran. Thus the distribution is not unlike 

 that of the tapirs, and we should look foi^ 

 fossil remains in the Tertiary beds of the 

 Palsearctic and Nearctic regions, were not the 

 preservation of Peripatus in this manner so 

 unlikely. The African species may possibly 

 he explained by an accidental passage (e. g., 

 on a floating tree) across the Atlantic, rather 

 than as indicating an ancient land-bridge, or 

 a long migration via Asia or Europe. It is 

 noteworthy, however, that in certain respects 

 it approaches the Malayan Eoperipatus, so 

 that it would be possible to regard the latter as 

 having reached Asia by way of Africa, the 

 Peripatus of the latter continent remaining as 

 a relic of a very ancient route of migration. 



The genus Peripatus, as now understood, is 

 divided into three groups or subgenera, as 

 follows : 



(1) Oroperipatv^ n.n. ; "Peripates andi- 

 coles." 12 species. Type P. lankesteri Bou- 

 vier. 



(2) Peripatus s.str. ; " Peripates caraibes." 

 17 species. Type P. juliformis Guilding. 



(3) Mesoperipatus Evans; "Peripates afri- 

 cains." 1 species. Type P. tholloni Bouvier. 



Professor Bouvier does not use any sub- 

 generic names; I have ventured to supply one 

 for the Andicolous group. This appears to 

 be the most primitive of the three, and might 

 better be. considered typical Peripatus, had not 

 the name been first applied to a Caribbean 

 species. 



The distribution of the two neotropical 

 groups is exceedingly interesting, one being 

 characteristic of the Andean chain, the other 

 of the more eastern regions and the West 

 Indies. Oroperipatus has three species in 

 Bolivia, six in Ecuador, one in Colombia, and 

 finally enters Mexico on the west side, getting 

 as far north as Tepic (P. eiseni Wheeler). 

 The subgenus Peripatus has nine species and 



subspecies in the West Indies, two in the 

 coast region of BrazU, five in the Guianas, 

 three in Venezuela, one each in Costa Eica 

 and Nicaragua, and reaches Mexico on the 

 east coast, exter-ding to Vera Cruz. If we 

 think of the two subgenera as reaching Cen- 

 tral America by way of the isthmus of Pan- 

 ama, it is not at first apparent why they should 

 have carefully kept to the same two sides of 

 the country they occupied in the southern 

 hemisphere. It is to be remarked, however, 

 that P. eiseni is a mountain species, occurring 

 at an altitude of 4,000 feet, and so its ances- 

 tors doubtless followed the Andean chain; 

 whereas the Caribbean types probably clung 

 to the coast line, and consequently to the east. 

 Whether through Tertiary time these animals 

 were widely distributed over North America, 

 or whether they reached that continent only at 

 the time when the isthmus rose above the sea 

 and permitted the ingress of various well- 

 known neotropical genera, must for the pres- 

 ent remain uncertain. 



The details of the distribution of the West 

 Indian species are worth investigating. There 

 are apparently no species in Cuba and the 

 Bahamas, and none have been described from 

 Hayti, though they may be expected there. 

 The lesser islands are populated by various 

 forms of the P. juliformis type, and Professor 

 Bouvier has shown that the Jamaican forms, 

 which I formerly regarded as variations of a 

 single species, really represent two, the so- 

 called P. jamaicensis swain-sonw Ckll. being in 

 fact a member of the juliformis series. It is 

 indeed so close to juliformis that Bouvier 

 makes it a variety of that species, though in 

 view of its isolation I should prefer to treat 

 it as a species, P. swainsonoe. It is practically 

 certain, no doubt, that P. swainsonce reached 

 Jamaica from the east, along the chain of 

 islands; but it seems probable that the other 

 species, P. jamaicensis, reached the island 

 (along with Capromys and other animals) 

 from Central America, by way of a land 

 bridge or chain of islands now represented 

 by large areas of shallow water. Professor 

 Bouvier, to whom I communicated this sug- 

 gestion, agrees with me that it is probably 

 correct; it has the advantage of accounting 



