May 2, 1902.] 



SCIENCE. 



691 



their descendants show in every case a 

 change in the proportionate diameters. As 

 an illustration, the common clam, Mya are- 

 nana, found in the shell heaps has a 

 greater height in proportion to its length 

 than that of the recent forms. This index 

 is higher north of Cape Cod, both in the 

 ancient and recent. The same differences 

 are found in Japan and in England. The 

 index is very much higher in the glacial 

 deposits. For this and other reasons the 

 change in the index is correlated with tem- 

 perature. 



It is furthermore pointed out that re- 

 lated forms in Japan and New England 

 had changed in precisely the same manner. 

 Full details with tables of measurements 

 will be given in the memoir on the subject. 



Professor Arnold E. Ortmann, of 

 Princeton, in a paper on the 'Distribution 

 of Fresh Water Decapods and its Bearing 

 upon Ancient Geography,' said that the 

 present and former distribution of these 

 animals suggested the following land con- 

 nections in former times : 



1. Northeastern Asia with northwestern 

 America, across Bering Sea. 



2. East Asia (Sinic Continent) with 

 Australia, over the Malaysian islands. 



3. South Asia and Africa, by way of 

 Madagascar. 



4. New Zealand with Australia, by way 

 of New Caledonia and New Guinea. 



5. Australia with Antarctica, and An- 

 tarctica vsdth South America (Archiplata). 



6. The Greater Antilles with Central 

 America. 



7. Africa with South America (Archi- 

 brazilia). 



Professor "W. M. Davis, of Cambridge, 

 in a paper on 'Systematic Geography,' said 

 that the accimiulation of an ever-increas- 

 ing store of facts under the broad subject 

 of geography makes it desirable to estab- 

 lish a classification with respect to which 

 the facts may be arranged, not only for the 



convenience of putting them away in good 

 order and of readily finding them again 

 when wanted, but even more for the sake 

 of the better understanding that comes 

 from association and correlation. Geo- 

 graphical classification may be by kinds or 

 by places, systematic or regional; but the 

 first should precede the second. The first 

 provides a scheme whereby all similar 

 items, whatever their place of occurrence, 

 may be brought together imder a single 

 category; the second describes all the items 

 of a certain region as examples of knovra 

 categories, and presents them in an order 

 that expresses their systematic relation- 

 ships. There is to-day no precise agree- 

 ment as to the total content of geography, 

 much less as to the subdivision and system- 

 atic arrangement of its parts; but if the 

 study of the earth in relation to its in- 

 habitants be taken as a sufficient definition 

 of the subject, its prime divisions must be 

 the physical environment of organisms and 

 the responses of the organisms to their en- 

 vironment. Each of these divisions is then 

 to be subdivided into many categories, and 

 each category is to be rationally described, 

 to be illustrated by typical examples, and 

 to be traced through its relationship to the 

 categoi'ies of the other prime divisions of 

 the subject. The innumerable relation- 

 ships thus disclosed constitute the subject 

 of geography proper, and it is as an aid to 

 their systematic treatment that the pro- 

 posed classification of the subject as a 

 whole is iindertaken. 



Mr. Henry G. Bryant, of Philadelphia, 

 in his paper entitled 'Drift Casks in the 

 Arctic Ocean,' gave the present status of 

 an experiment worked up by Admiral 

 George W. Melville, U. S. N., and himself, 

 Avhich aimed to test the speed and direction 

 of Arctic currents by means of a series of 

 drift casks set adrift in the Arctic Sea 

 north of Alaska. He called attention to 

 the fact that the scheme was the outcome 



