VL] ON THE METHOD OF ZADIG. 147 



that, as we trace them back in time, their ancestors 

 gradually cease to exhibit those special modifications 

 which at present characterise the type, and more nearly 

 embody the general plan of the group to which they 

 belong. 



Thus, in the well-known case of the horse, the 

 toes which are suppressed in the living horse are found 

 to be more and more complete in the older members 

 of the group, until, at the bottom of the Tertiary 

 series of America, we find an equine animal which has 

 four toes in front and three behind. No remains of 

 the horse tribe are at present known from any Mesozoic 

 deposit. Yet who can doubt that, whenever a suffi- 

 ciently extensive series of lacustrine and fluviatile 

 beds of that age becomes known, the lineage which 

 has been traced thus far will be continued by equine 

 quadrupeds with an increasing number of digits, until 

 the horse type merges in the five-toed form towards 

 which these gradations point ? 



But the argument which holds good for the horse, 

 holds good, not only for all mammals, but for the 

 whole animal world. And as the study of the pedi- 

 grees, or lines of evolution, to which, at present, we 

 have access, brings to light, as it assuredly will do, 

 the laws of that process, we shall be able to reason 

 from the facts with which the geological record 

 furnishes us to those which have hitherto remained, 

 and many of which, perhaps, may for ever remain, 

 hidden. The same method of reasoning which enables 

 us, when furnished with a fragment of an extinct 

 animal, to prophesy the character which the whole 



