A PROBLEM IN EVOLUTION 427 



some animals that were intermediate in structure between them, and 

 which made their appearance at some time not later than the Silurian 

 period. The ostracoderms were the only ones known to science in any- 

 way likely to fulfil these requirements; it was, therefore, of the utmost 

 importance to learn something more about these mysterious, extinct 

 animals, for they appeared to contain the solution of the whole problem. 



The first opportunity to test this side of the problem came in 1900- 

 01, with a half-year's leave of absence from college duties. The pros- 

 pects of success, however, were very small, especially for one trained as 

 a laboratory morphologist and embryologist, and without experience in 

 field geology or in paleontology. There was a bare chance that a 

 reexamination from another point of view of the material preserved 

 in the British Museum and other institutions, and that had already 

 been studied by Huxley, Lankester, Traquair, Woodward and others, 

 might reveal some suggestive details overlooked by these past masters 

 of the subject. Failing that, it would be necessary to go into the 

 field and dig up new material that, to serve our purpose, would have 

 to be more perfectly preserved than any that had, by chance, been found 

 in the preceding three quarters of a century. It was not a promising- 

 outlook, but the opportunity was gladly accepted. 



We first visited the great museums of England and Scotland, and 

 the localities where, in the early thirties, Hugh Miller unearthed the 

 first specimens of these animals known to science, which he afterward 

 described with such remarkable literary skill and enthusiasm in his 

 " Footprints of the Creator," and in the "Old Bed Sandstone." 



But the best material available in England and Scotland was pro- 

 vokingly incomplete in regard to the very structures it was most impor- 

 tant for us to know about. However, some unexpected and suggestive 

 details were found that greatly added to the already keen excitement 

 of the search (Fig. 4). 



It was then decided to visit the famous Silurian quarries on the 

 Island of Oesel, in the Baltic Sea, and the museums of St. Petersburg 

 and Moscow, where many of the Oesel fossils were preserved. 



The representatives of the ostracoderms (Tremataspis) found in 

 the Silurian rocks of the Island of Oesel are only about three inches 

 long. But in spite of their small size, they are, in some respects, 

 admirably preserved in a soft, fine-grained limestone; and they prom- 

 ised to yield important data. In fact, the specimens that were obtained 

 there showed the presence of jointed appendages and shell-covered, 

 stalk-like eyes. These structures were unlike those of true fishes and 

 more like what one would expect to find in some free-swimming sea- 

 scorpion. In that respect the results were highly satisfactory, and 

 added still more evidence in favor of our first supposition (Fig. 5). 



