36 PROFESSOR A. M. MARSHALL. 



ancestors of smaller size, but there is also much evidence to show that 

 increase in size beyond certain limits is disadvantageous, and may lead 

 to destruction rather than to survival. It has happened more than 

 once in the history of the world, and in more than one group of 

 animals, that gigantic stature has been attained immediately before 

 extinction of the group, a final and tremendous effort to secure 

 survival, but a despairing and unsuccessful one. The Ichthyosauri, 

 Plesiosauri, and other extinct reptilian groups, the Moas, and the huge 

 extinct Edentates, are well-known examples, to which before long will 

 be added the elephants and the whales, and, it may be, ironclads 

 as well. 



The whole question of the influence of size is of the greatest 

 possible interest and importance, and it is greatly to be hoped that it 

 will not be permitted to remain in its present uncertain and un- 

 satisfactory condition. 



It may be suggested that Amphioxus is an animal which has under- 

 gone reduction in size, and that its structural simplicity may, like that 

 of Limapontia, be due, in part at least, to this reduction. Such 

 evidence as we have tells against this suggestion ; the first system to 

 undergo degeneration in consequence of a reduction in size is the 

 respiratory, and the respiratory organs of Amphioxus, though very 

 simple, are also, for a vertebrate, unusually extensive. 



We have now considered the more important of the influences 

 which are recognised as affecting developmental history in such a way 

 as to render the recapitulation of ancestral stages less complete than 

 it might otherwise be, which tend to prevent ontogeny from correctly 

 repeating the phylogenetic history. It may at this point reasonably 

 be asked whether there is any way of distinguishing the palingenetic 

 history from the later cenogenetic modifications grafted on to it ; any 

 test by which we can determine whether a given larval character is or 

 is not ancestral. 



Most assuredly there is no one rule, no single test, that will apply in 

 all cases ; but there are certain considerations which will help us, and 

 which should be kept in view. 



A character that is of general occurrence among the members of a 

 group, both high and low, may reasonably be regarded as having 

 strong claims to ancestral rank ; claims that are greatly strengthened 



