264 DISCOVERY REPORTS 



Rubrius subfasciotus Sim. appears to be common in Tierra del Fuego, and has 

 previously been collected on Hermite Island. 



Coming now to the third and new species, which I am calling Myro frigida, we 

 find much of interest in its discovery in South Georgia. 



No spiders had previously been collected in South Georgia and now, out of sixteen 

 specimens from four or five difl^erent localities, this is the only species to be found. It is 

 small and it is not conspicuous, so we have grounds for thinking that it is the only spider 

 inhabitant of South Georgia. An interesting fact, which incidentally supports this view, 

 is that the isolated Antarctic islands of Kerguelen and Macquarie also have one spider in- 

 habitant in each case peculiar to themselves and in each case belonging to the genus Myro . 



So far as I am aware the distribution of this genus is as follows: Cape of Good Hope, 

 Snares Island, Macquarie Island, Tasmania, Kerguelen and now South Georgia.^ 

 Each locality has its own peculiar species and Snares Island can boast of two. 



Myro, and also Riibrius, belong to a sub-family of the Agelenidae, the Cybaeinae, 

 and L. Berland has recently pointed out that no less than eight out of the fourteen 

 species known from Bounty, Snares, Auckland, Campbell, Macquarie and Kerguelen 

 Islands belong to this sub-family,^ or nine to the family Agelenidae. 



Thus by adding South Georgia to the above list of islands we can say that of their 

 known fauna ten out of fifteen, or two-thirds, belong to the family Agelenidae and nine 

 out of fifteen, or three-fifths, to the sub-family Cybaeinae. Furthermore, of the nine 

 Cybaeinae seven are peculiar to the particular island on which each is found, the other 

 two being shared by Auckland and Campbell. 



This is a very different state of affairs to that existing in the Arctic. There the Liny- 

 phiidae are the predominant family, and there the isolated small islands do not have 

 any, or at any rate many, species peculiar to their faunas. Taking the four islands of 

 Jan Mayen, Bear Island, Spitsbergen and the New Siberian Islands we find that out of 

 ten species, nine belong to the Linyphiidae and one, Micaria eltoni jacks., to the Clubi- 

 onidae. Three of the four Jan Mayen species are found on Spitsbergen and the fourth 

 on Bear Island. The three species recorded for the New Siberian Islands are all present 

 on Spitsbergen. 



This comparison of Arctic with Antarctic islands shows us (i) that in the former it is 

 the Linyphiidae, in the latter the Agelenidae that are dominant, and (2) that in the 

 former the restriction of species to particular islands is the exception, whilst in the 

 latter it is the general rule. Have these two points any connection with one another? Is 

 there any connection between the facts that in the Arctic eight out of nine Linyphiids on 

 these four islands are not peculiar to any one of them (and most of them are found in 

 Arctic Europe or Greenland), whilst in the Antarctic seven out of nine Cybaeinae are 

 restricted to particular islands? Elsewhere^ I have stressed the fact that the ballooning 



1 Myro backhaiiseni Sim. and M. paupercida Sim. described from the south of South America have now 

 been transferred to the genus Myropsis and the family Dictynidae. 



^ Compte Rendu Sommaire des Seances de la Soc. de Biogeographie, 1930, p. 90. 

 * Proc. Zool. Soc. Pt 4, 1929, p. 633. 



