president's address. 29 



conditions ; (3) the stricter the environments of a type are, the less 

 are its means of dispersal ; and (4) the greater the facilities of a type 

 are for extension of its range, the wider should be its range. 



Now, these propositions are such that they might at first appear 

 to be shallow truisms', but the emphatic proviso that all other Ihmgs 

 should be equal is especially necessary to keep us from error in ap- 

 plying the several propositions to the actual faunas. A brief glance 

 at the past history of some classes will render the necessity evident. 

 For example, on the one hand, even in the Palaeozoic period, 

 among others, representatives of the classes of Lamellibranchiates, 

 Gastropods, Arachnids, and Insects lived, which are inseparable 

 from families still in existence, while, on the other hand, extremely 

 few, if any, families of mammals or birds have a history which ex- 

 tends back as such into the Eocene Tertiary ; between the two cate- 

 gories expressed by such types the fishes form an intervening term. 

 Presumably, we should, therefore, be justified in expecting a wide 

 distribution of the first series of types, and, inversely, from their 

 geographical distribution, we should expect a long life history for 

 them. Our provisions in both cases would be justified by the facts. 

 And further, a study of the distribution of those types shows that 

 their ranges, in higher groups, are very imperfectly co-ordinate with 

 the present distribution of land. 



The fresh-water fishes had become more differentiated than the 

 invertebrate types, in Mesozoic and Tertiary times, and are, there- 

 fore, better exponents of the later past relations of continental 

 areas than they are. Their presence, as well as their absence, are 

 very significant, and the bearings of such distribution have been 

 partly indicated in connection with the proposition to recognize 

 the Eogaean and Cenogsean worlds. Their study will avail much 

 hereafter in the reconstruction of the continents of the older epochs, 

 and the solution of the derivations of their faunas. Suffice it now 

 to express the opinion that the fishes are among the best indicators 

 of the ancient continental ar^as, and that the relations of the sev- 

 eral southern continents as to their ichthyic faunas can be best under- 



