310 A. H. Clark. 



wliich shows periods of greater abimdanco or nodes, and these nodes are 

 at approximately regulär intervals. 



The intervals between these nodes arrange themselves according to 

 the ascending powers of 2. If we take 50 fathoms as cur unit (a) and 

 calculate the nodes we would find nodes indicated at 50 fathoms (a), at 

 200 fathoms ('2^a), at 400 fathoms {2'^a), at 800 fathoms {-I^a] and at 

 ] 600 fathoms (2'^a). We actually find nodes at 0—50, 100—150, 300—400, 

 750 — 900, and 1600 — 2000 fathoms. The spacing of the nodes is appro- 

 ximately represented by the Inversion of the curve representing the de- 

 crease in temperature with depth. 



In the Caribbean area the same thing occurs, biit the nodes are 

 closer together, being represented by the series a, 2 a, '2^a, 2^a and 2*a. 

 This is probably the result of the much greater bathymetrical ränge of the 

 individual speeies. 



This path of migration, from southeastern Africa to the Antillean 

 region, has long been closed, and it is a significant faet that the speeies 

 which entered the Atlantic by this route are largely confined to the Ameri- 

 can side, being represented on the opposite coast by but a single form, 

 which is capable of existing in deep water. 



This suggests that the inflvix of eastern speeies into the Atlantic by 

 this route was along the southern border of a large land mass which in- 

 cluded both Madagascar and the West Indian region, but which did not 

 include the present south Africa. The southerly extension of Africa into a 

 region of comparatively cold water (or the closing of a route once 

 open across south central Africa) interposed a barrier to the further 

 spread of East Indian forms into the Atlantic, while the subsidence of the 

 land in the Atlantic area left the intrusive speeies segregated in the An- 

 tillean region and on the southwest African coast. Unfavourable conditions 

 in the latter locality resulted in the disappearance from this region of all 

 of these types except one which, able to live at a considerable depth as 

 well as near the surface, has succeeded in surviving. In the Antillean region 

 they have been able to persist throughout the succeeding ages and to 

 survive all the geological changes by which the geography of the region 

 has been fundamentally altered, and we therefore find in the West Indies 

 a crinoid fauna strikingly like that of Madagascar, the Mascarene Islands, 

 and the adjacent portion of Africa, in correlation with the well known 

 similarity between the land fauna of the two districts. 



From the evidence aftorded by the stalked crinoids of the West 

 Indies, belonging to genera not represented on the west coast of Africa 



