278 NATURAL SCIENCE. April, 



method of development, yet not the most sanguine of them will hope 

 to find so perfect an epitome of phylogeny in the majority of cases. 

 The necessary compression, the constant elimination of unnecessary 

 stages, the modifications required by larval conditions — all these things 

 make it truly remarkable that we should get as much recapitulation 

 as we do. Yes, we do get it, and this is " the way in which it does 

 actually occur," let Mr. Hurst deny it as much as he pleases. In fact, 

 no more perfect example of this impossible and non-existent method 

 of development could be imagined than that which is actually afforded 

 us by the palaeontological history of the Ammonites. Does Mr. Hurst 

 dispute the reality of the facts made known to us by Wiirtenberger, 

 Waagen, Branco, Hyatt, Buckman, and many others ? Not everyone 

 will accept the alterations in classification proposed by these workers, 

 but no one has hitherto been found to deny their facts. The reason 

 why no one has done so is obvious : anyone can verify them for him- 

 self. It has been shown over and over again, by workers approaching 

 the subject quite independently and from different points of view, that 

 the adolescent stage of any species of Ammonite resembles the adult 

 stage of its immediate ancestor, and that the larval stage resembles 

 the adult of a previous ancestor ; while, on the other hand, the senile 

 stages of the same species foreshadow the features of the adult 

 offspring. These facts have not merely been worked out by students 

 in museums, by dissection of well-preserved individuals, and com- 

 parison of large quantities of specimens, but have been confirmed by 

 minute labours in the rocks themselves, and by the careful tracing of 

 the species from zone to zone over large tracts of country. 



A suitable example is presented by the descent of Coroniceras 

 trigonatum from a smooth ancestor, as indicated in Hyatt's " Genesis 

 of the Arietidae," Summary-plate xii. Hyatt here gives the ancestry 

 as follows :— C. trigonatum, C. gmiiendense, C. lyra, C. rotifornie, 

 C. saiizeanmn, C. hridion, Arnioceras kridioidcs, A. semkostatum, and 

 Psiloceras planorhe, var. Icve. To obtain an independent opinion, I 

 applied to my friend, Mr, S. S. Buckman, who informs me that he 

 agrees with the first six names except as to the interposition of 

 C. sauzcanum. He, however, " would not like to say that the species 

 oi Arnioceras are the actual ancestors ; but they are the morphological 

 equivalents undoubtedly of those ancestors." There is also consider- 

 able objection to taking a retrogressive type like Psiloceras planorhe as 

 the starting point of a new series. At the same time, even those who 

 refuse to regard this particular species in the same light as Professor 

 Hyatt, w^ill admit that the ancestor must have been a somewhat 

 similar smooth, keelless, and round-sectioned form ; in fact, Buckman 

 favours Arnioceras miserabilc, which Hyatt himself gives as an 

 ancestral form. Accepting, then, this line of descent, we may draw 

 up the accompanying table, which traces the development of three 

 main characters in both ontogeny and phylogeny. The first column 

 describes the infantine stage, the second ihe adolescent, and the 



