300 F. H. EDGEWORTH. 



or myotomes comes last. This general rule leads to hesitation 

 in accepting the existence of anterior nerve rot)ts Avithout 

 corresponding somites or myotomes as evidence of assimi- 

 lated segments, e.g. deductions from the observations of 

 Chiarugi and Martin in Mammals. 



The theory of Fiirbringer is based on the probability of the 

 primitive nature of the conditions found in Selachians. But 

 from the foregoing table oi: the observed number of assimi- 

 lated spinal segments in various Vertebrates it would appear 

 that the descriptive adjectives applied to some neocrania are 

 not deserved. As determinated by the number of assimilated 

 spinal segments the Amphil>ian neocranium is shorter than 

 that of Selachians. It was therefore maintained that the 

 occipital region of Amphibians corresponds to amultiplum of 

 spinal segments. The difficulty of doing so is emphasised by 

 the absence of any direct evidence in its favour. If the 

 muscles of the head in Amphibians and Selachians be com- 

 pared it is clear that the condition in the former is far more 

 primitive than in the latter, and that many cranial muscles 

 of Scy Ilium pass through what may be regarded as an 

 Amphibian stage during development ; and if the observed 

 facts in regard to the number of assimilated spinal segments 

 be taken sans parti pris the condition of the skull tells 

 the same tale. Fiirbringer states that the junction of the 

 skull and vertebral column is at the same place in Sauropsida 

 and Mammalia; hence the live occipital nerves in Eeptilian 

 embryos are called v, w, x, y, z ; and the three in Mammals 

 X, y, z, so that the last assimilated nerve is the same — z. 

 But in Mammals there appear to be only three assimilated 

 somites, in Keptiles four or five. The argument drawn from 

 the existence of a pro-atlas is probably of no great weight in 

 determining the limits of the skull and vertebral column, for 

 in Sphenodon (loc. cit.) that structure is the persisting costal 

 process of the last coalescing vertebra, and the same may be 

 true in Mammals without there being any but a serial 

 homology between these last coalescing vertebra). 



The conclusion which might be drawn from the number of 



