Herrick, Criteria of Homology. 205 



in a single trunk, the homology is immediately disturbed. For, 

 not only does the foreign admixture come from a different cerebral 

 segment, but it is of totally different functional composition, con- 

 necting with a different type of sense organ peripherally and with 

 a different cerebral coordination system. 



A survey of the phylogenetic history of the facialis nerve pre- 

 sents constantly recurring phases of the problem. This nerve was 

 in primitive vertebrates a branchiomeric nerve, supplying a gill- 

 bearing segment and containing at least four components. The 

 stages in its metamorphosis into a nerve supplying chiefly the super- 

 ficial mimetic facial musculature of man can be clearly read by 

 the comparative anatomist. 



In the course of this phylogeny some new components are added, 

 some are totally lost and the survivors experience manifold changes 

 of function and rearrangement of rami. Though the identity of 

 .the nerve as a segmental unit is never lost, yet its perfect homology 

 throughout the series is certainly open to discussion ; and the mor- 

 phological position of some of the rami whose composition varies 

 from type to type by reason of peripheral anastomoses with other 

 segmental nerves, such as the trigeminus and glossopharyngeus, is still 

 more ambiguous. 



I would suggest as a basis for further discussion the following 

 rules for the fixing of homologies in the peripheral nervous system 

 of vertebrates : 



(1) If a nerve is a member of a meristic series (cranial or 

 spinal), the preservation of its individuality in the comparative 



. series of animals requires that its roots must come from the same 

 segment or segments throughout the phylogenetic series. If there 

 has been a shifting of one or more roots to another segment the 

 homology is thereby to that extent impaired. 



(2) But if the nerve in question is a composite of roots from 

 several segments (like the hypoglossus), the individuality of the 

 nerve is not necessarily destroyed by a shifting of the whole series 

 forward or backward, or by the inclusion of more or less segments 

 of the same kind in the complex, provided the relations to adja- 

 cent segmental nerves are not fundamentallv altered. Neverthe- 



