1902.] NATURAL SCIENCES OF PHILADELPHIA. 231 



structural couditious of this stage control, more especially where 

 the ontogeny is Avithout metamorphosis, the nature of the condi- 

 tions that precede it. Metamorphosis and larval conditions have 

 probably been secondarily acquired, and direct development is the 

 more primitive; for metamorphosis is a result of change of habit 

 at different periods, and such change of habit is not as primitive as 

 a habit continuous through the life of the individual, since it is 

 more complex. It will, of course, be understood that in the case 

 of certain phyla, such as the Annelida, I grant that the larval 

 type of development may be primitive, and that forms which do 

 not show it have lost it ; but that it could not have been primitive 

 for the ancestors of the Annelida. There is also good reason to 

 conclude that the larvse of different phyla of organisms may be 

 alike, because they are ontogenetic forms of essentially the same 

 mode of life, and because their organs are comparatively few in 

 number and at an early stage of specialization. The earlier the 

 stage of the ontogeny, the greater in general should be the agree- 

 ments in structure, not because the homologies are more comprehen- 

 sive at that time, but because there are fewer structural parts to 

 show differences. 



Thus far we have criticised the application of the biogenetic 

 theory to the determination of homology. These criticisms might 

 be multiplied and examined more in detail; and they are objec- 

 tions which have been made before, though not from the standpoint 

 of the necessity of classifying the organism at its terminal stage. 



Now we have to consider in what respects the study of onto- 

 genetic conditions is necessary in the search for homologies. 

 Homology is a likeness of descent, is homophyly, whether it be 

 based upon adult or embryonic structure, provided that there is a 

 good basis employed in its determination. It is an erroneous posi- 

 tion to hold that the adult anatomy alone is sufficient for deter- 

 mining all homologies; it is equally erroneous to hold that homo- 

 logies can be determined on the basis of similarity in ontogenetic 

 formation alone. Whenever a structure is greatly degenerate in 

 the adult condition, a study of its preceding ontogenetic stages is 

 often the only guide we have as to its homologies."* Then in organ- 



" Often, however, a highly degenerate structure can be explained by 

 comparative anatomy as well as by embryology. Thus the pineal gland 

 of a bird and the pineal eye of certain reptiles. 



