COOKE: ON GEOGRAPHICAL DISTKIBUTION. 107 



example of a district on which we are remarkably destitute of 

 information we maj' take the whole South American coastline from 

 Venezuela to Buenos Aires, and even further south. Wliat little we 

 know inclines us to believe that the Mollusca of these thousands 

 of miles of coast is typically Antillean in character; at any rate 

 Purpura hcemastovia, L., both typical and in varieties, is found as 

 far south as Kio Grande do Sul (32° S.) and the mouth of the lUo 

 de la Plata. That many thoroughly littoral species should be able 

 to cross the present volume of the discharge of Amazonian fresli- 

 water, covering some hundreds of square miles, seems so incredible 

 that it may be held that the existing coast fauna antedates the 

 existence of that and other streams, at least in their present 

 immensity.' 



Verrill has pointed out" that the entrance of Long Sound and the 

 bays and sounds lying south of South Massachusetts are inhabited by 

 two separate moUuscan faunas, the shnllower waters of tlie bays being 

 occupied ciiiefly by southern forms belonging to what he then calls 

 the Virginian fauna, while the deeper channels of the central parts 

 of the sound are inhabited exclusively by a northern fauna. The 

 cause of this apparently anomalous state of things is that an offshoot 

 of the cold Arctic current which sweeps round Nova Scotia sets into 

 the middle of the sound and produces, both at the surface and at the 

 bottom, a change of temperature, which, within a space of only 

 2 miles, amounts to as much as 5° F. Thus the littoral fauna is of 

 a comparatively southern type, while even the shallow-water fauna, 

 at depths of no more than 18 to 39 fathoms, is strictly northern, 

 consisting of the following amongst other species : Molgida pihdaris, 

 Glandula mollis, Cardita borealts, C. novanglia, Yoldia sapotilla, 

 Y. limatula, JVucula proxima, Astarte quadrans, A. castanea, Modiolaria 

 niqra, 31. corrugata, Chrysodomus pygmceaj Margarita obscura, Cylichna 

 alba, and many others. 



Much useful aid in exploration may be gained from geology in 

 showing that certain modifications of climate and of elevation, 

 otherwise imsuspected, must have taken place. Thus, to take one 

 instance out of many, G. Bardason has shown, ^ from the evidence of 

 Pleistocene marine beds in North Iceland, that within comparatively 

 recent times the sea was at least 4 metres above its present level, 

 with the effect that the temperature of that particular region must 

 have been higher than it is at the present epoch, or mucli as it is now 

 in South- West Iceland. This is shown by the presence in the 

 deposits of Purpura lapillus and Zirphcea crispafa, and by the absence 

 of Pecten islandicus. As the sea retreats the temperature, in northern 

 regions, becomes lower, and the conditions assume a more Arctic 

 character. 



' W. H. Dall, " Additional Notes from the Coast of Southern Brazil " : 

 Nautilus, vol. vi, pp. 109-12, 1893. "List of Shells collected at Bahia, 

 Brazil, by Dr. H. von Ihering " : ibid., vol. x, pp. 121-3, 1897. 



^ Ann. Mag. Nat. Hist., ser. IV, vol. ix, pp. 92-7, 1872. 



■* " Maerker efter Klima- og Niveanforandringer ved Hunafloi i Nord-Island " : 

 Vid. Medd. Copenhagen, 1910 (ii), pp. 35-79. 



VOL. XI.— JUNE, 1914. 8 



