Febeuaey 28, 1896.] 



SCIENCE. 



319 



represented in all three of the present 

 southern continents, or in even two of them, 

 the case would be different. The single 

 order Passeres includes families peculiar 

 respectively to South America, Africa and 

 Austi-alia, which are far more closely re- 

 lated to each other than are the several 

 orders of the Ratitse inter se; yet no one 

 thinks of urging these Passerine groups as 

 evidence of a former Antarctic continent. 

 They are supposed to have originated in- 

 dependently where they are now found and 

 to have never existed elsev/here. 



There are, on the other hand, several 

 families of Carinate birds, belonging to dif- 

 ferent orders, which inhabit the tropi- 

 cal and subtropical regions of both the 

 Eastern and Western hemispheres, but 

 which now aud for long ages past have had 

 no possible means of migration from Amer- 

 ica to Africa, or to India, or to Australia. 

 That the present ISTew "World and Old 

 World representatives of these several 

 groups must have had, respectively, a com- 

 mon origin is beyond question ; and it is be- 

 Keved to be equallj^ beyond question that 

 they reached their present areas of distribu- 

 tion by the northern land route that formed 

 the means of intercommunication between 

 the northern land masses for so many of the 

 widely dispersed terrestrial forms of life. 



Another factor bearing on the general 

 question is the early origin of many of the 

 existing genera of birds, most of the known 

 Pliocene genera still surviving, while many 

 of the Lower Miocene and Upper Eocene 

 genera of Europe and North America are 

 in some cases identical, in others closely al- 

 lied, to genera stiU living. Some of them 

 are now restricted to the tropics, but their 

 ranges formerly extended far to the north- 

 ward of their present limits. 



In short, birds afford no clear evidence 

 in favor of the existence of a former Ant- 

 arctic continent, and mammals only that af- 

 forded by the distribution of the Marsupials. 



Vertebrata of the Sea. 'Bj Theo. Gill, Wash- 

 ington. 



On account of the enforced absence of Dr. 

 Goode, detained in Washington by official 

 business, and at his request. Dr. Gill con- 

 sidered the subject assigned to him — the 

 fishes of the sea in relation to the Antarctic 

 continent. 



There is really no direct evidence fur- 

 nished hy sea fishes bearing on the question 

 at issue. There are, however, some facts 

 which may throw light on a certain phase 

 of the question. The fishes of the Antarc- 

 tic seas are very imperfectly known, but 

 the few that are known are of much interest 

 and belong to two very distinct categories. 

 On one hand, we have a few species be- 

 longing to a couple of families only occur- 

 ring in the extremely cold waters — the 

 Chsenichthyids and Harpagiferids. The 

 genera of these families have been referred to 

 the family of Trachinids, but really mani- 

 fest no affinity to the typical forms of that 

 group. The only inference that appears to 

 be derivable from the two families is that 

 the supposititious Antarctic continent may 

 have been in all Tertiary geological times 

 at least deeply indented by extensions of 

 the ocean far towards the Pole. 



On the other hand, in the Antarctic seas 

 recur representatives of genera which have 

 been only found in high northern waters, 

 such as Myxine, Squalus, and Merlucius, and 

 those representatives are so closely related 

 as to have been united in two cases as con- 

 specific. It appears to be most reasonable 

 to postulate for such types derivation from 

 a common source, and that their extension 

 may have been effected in the cold waters 

 of the ocean depths. It is more than pos- 

 sible that, under favorable conditions, spe- 

 cies of Myxine, Squalus and Merlucius may 

 yet be found in the cold deep waters below 

 even equatorial seas, for it is to be remem- 

 bered that all have an extensive bathymet- 

 rical range. 



