13G 



AXXALs XEW YORK ACADEMY OF SCIEXCES 



In addition to. and partly because of, our relative ignorance of ancient 

 spiders, we known but little of araneid phylogenv. 'SVe do know that 

 spiders were already fairly well distributed, at least in the northern 

 hemisphere, in the Carboniferous. Modern families and even genera 

 were well differentiated in Oligocene. Therefore, as far as distribution 

 problems are concerned, the present-day distribution of primitire genera 

 in a given family or even primitive species in a given genus is of as 

 much, or more, importance as the present-day distribution of primitive 

 families. We unfortunatelv lack a sufficient knowledge of the compara- 



A, 



Fig. 5. — Distribution of Archceidce 

 Fossil (amber 1 Arch a: a ; B. Living Archaa ; C. M€<ysmaiichenius. 



tive anatomy of sjiiders to enable us to use such data in the solution of 

 the problems of distribution. In fact, widely discontinuous distribution 

 is, at present, the best indication we have of relative antiquity and its 

 use is rendered hazardous by reason of the possibility- of polyphyletic 

 origins — a possibility which has been rendered more probable by recent 

 work in experimental evolution. 



An analysis of the distribution of the world's spider fauna would be 

 out of place in this paper and only a few points which seem to have a 

 bearing on the West Indian problem will be taken up. 



The family Archieidse is an interesting one in this connection (see 

 figure 5) since Baltic amber cont-ains a 2"enus (Archcea) of which the 



