204 PRIMARY FACTORS OF ORGANIC EVOLUTION. 



can perform no other function, and soon die, while in 

 some Diptera the incomplete larvae themselves repro- 

 duce, so that the metamorphosis is never completed. 



This history is parallel to that proposed by Dohrn 

 to account for the origin of the Ammocoetes larval 

 stage of the Marsipobranchii. He supposes this form 

 to be more degenerate than the corresponding stage of 

 its probable ancestral type in the ancestral line of the 

 Vertebrata. An inactive life in mud is supposed by 

 Dohrn to have been the effective cause. An inactive 

 life on the leaves of plants, or in dead carcases, has 

 probably been the cause of the same phenomenon in 

 the Lepidoptera and Diptera. 



Thus we have developed an ontogeny within an 

 ontogeny, and a phylogeny within a phylogeny. These 

 facts do not, however, affect the general result in the 

 least. They only show us that the persistent larvae 

 of those animals which possess them, have a history of 

 their own, subject to the same laws of evolution as 

 the adults. It results that in many cases the phy- 

 logeny can only be determined by the discovery and 

 investigation of the ancestors themselves, as they are 

 preserved in the crust of the earth. In all cases this 

 discover}' confirms and establishes such definite con- 

 clusions as may be derived from embryology. It is 

 also clear that on the discovery of phylogenetic series 

 it becomes at once possible to determine the nature of 

 defective types. It becomes possible to ascertain 

 whether their rudimental parts represent the begin- 

 nings of organs, or whether they are the result of a 

 process of degeneration of organs once well devel- 

 oped. 



An excellent illustration of inexact parallelism is 

 to be found on comparison of man with the lower Ver- 



