236 FROM FISH TO PHILOSOPHER 



Iddney was an evolutionary adaptation to fresh water 

 and, with E. K. Marshall, Jr. {37} presented extended 

 evidence on this point, derived from the comparative 

 anatomy and physiology of the vertebrate kidney; and, 

 in 1932, in reviewing the problem of water regulation 

 in the teleost and elasmobranch fishes, he advanced new 

 physiological evidence in support of the fresh-water 

 theory {47}. 



Paleontologists who remained unconvinced of the cor- 

 rectness of Chamberlain's theory were favorably influ- 

 enced when, in 1935, A. S. Romer, one of America's out- 

 standing paleontologists, and his colleague, B. H. Grove, 

 presented a detailed re-examination of the habitat prob- 

 lem {45}. Romer and Grove considered in extenso only 

 the records of North America up to the close of the 

 Devonian period, but they included brief resumes of the 

 evidence from Europe and elsewhere, and their conclu- 

 sions were entirely in favor of the fresh-water theory. 



This theory has again been challenged, however, by 

 J. D. Robertson {39} of the University of Glasgow, and 

 R. H. Denison {31} of the Chicago Natural History 

 Museum. Robertson rehes heavily on the voluminous 

 literature, much of it of an older date, dealing with the 

 Ordovician and Silurian forms, to which he adds the well 

 known argument that the three Hving protochordate 

 groups, Hemichordata, Urochordata and Cephalochor- 

 data, as well as the Myxinoidea, are marine. More cogent, 

 perhaps, are the criticisms of Denison, who has sifted all 

 the available evidence and who concurs in the theory of 

 a marine habitat for the early vertebrates. Romer {43, 

 44}, though still emphasizing the role of the armor as 

 protection against the eurypterids, continues to adhere 

 to the fresh-water theory. 



To these weighty opinions I can add little. My pale- 

 ontologist friends point out that the problem is a most 

 difficult one, perhaps subject only to a Scottish verdict, 

 not proved either way,' and that the physiologist will 

 have to proceed on his own at the present time. As a 



