THE WONDERS OF LIFE 



to look on them with a certain distrust; some contest 

 their scientific value altogether, and say that they are 

 nothing but airy and untenable speculations. This is 

 especially the case with many physiologists who look 

 upon experiment as the only exact method of investiga- 

 tion, and many embryologists who think their sole task 

 is description. In view of these sceptical strictures, we 

 may recall the history and the nature of geology. No 

 one now questions the great importance and the various 

 uses of this science, although in it there is no possibility 

 of directly observing the historical processes as a rule. 

 No scientist now doubts that the three vast successive 

 formations of the Mesozoic Period — the Triassic, Jurassic, 

 and Cretaceous — have been formed from sea-deposits 

 (lime, sandstone, and clay), though no one was a witness 

 to the actual formation; no one doubts to-day that the 

 fossil skeletons of fishes and reptiles which we find in 

 these groups are not mysterious freaks of nature, but the 

 remains of extinct fishes and reptiles that lived on the 

 earth during those millions of years long ago. And 

 when comparative anatomy shows us the genealogical 

 connection of these related forms, and phylogeny (with 

 the aid of ontogeny) constructs their ancestral trees, 

 their historical hypotheses are just as sound and reliable 

 as those of geology ; the only difference is that the latter 

 are much simpler, and thus easier to construct. Phylog- 

 eny and geology are, in the nature of the case, historical 

 sciences. 



Hypotheses are necessary in phylogeny and geolog}^ 

 where the empirical evidence is incomplete, as in every 

 other historical science. It is no detraction from the 

 value of these to urge that they are sometimes weak and 

 have to be replaced by better and stronger ones. A 

 weak hypothesis is always better than none. We must, 

 therefore, protest against the foolish dread of hypotheses 

 which is urged against our phylogenetic methods by the 



37S 



