[35] THE OSTEOLOGY OF AMI A CALVA. 781 



tion with ganglionic enlargements at the distal extremity of the olfac. 

 troy nervelets do not exist. 



Besides, there are — though very rare — intermediate forms known be- 

 tween the two types we have indicated among fishes ; cases, for instance, 

 where the bulbus is placed half-way between the brain and the olfactory 

 membrane, and where it is connected with the former by a thin, soft 

 tractus ; with the latter by a strong, firm nerve at least four times as 

 thick. The only other case of this kind known up to the present time 

 has been noticed by Stannius in the Gadus raniceps fuscus ; and I 

 find quite a similar condition in the Oharacinidse, as in Hydrocyon and 

 Alestes. 



A mere superficial examination of these two types does not furnish 

 us with sufficient data to judge from, and decide which is the primary 

 form and which is the derived one. As in so many other cases, the ques- 

 tion can only be decided by the systematic — based upon other con- 

 ditions of organization — position of the forms that belong to one 

 or the other type. We now find that the first type occurs in all Se- 

 lachians, in Holocephals, and certain of the Teleostean groups, long 

 known to us as the primitive forms, as in the Siluroids, the Oyprinoids, 

 the GadidsB, and, as I have found, also in the Mormyridse. 



The second type is extensively found in the Ganoids and in the great 

 majority of the Teleosteans. With all this before us, no doubt can re- 

 main that the first type is the primitive one, and that from it the other 

 type has developed by a gradual shortening of the tractus and a length- 

 ening out of the nerve. 



It appears that in the Teleosteans the development of the olfactory 

 nerve is always brought about in the same way and with a uniform result. 

 The enlargement of the orbits leads to a fenestration of the lateral or- 

 bital wall at its anterior angle near where the bulbus olfactorius was 

 originally located, as one can see very well in the Characinides ; this 

 development extending further gives rise to an olfactory nerve, which 

 must of necessity pass through the orbits. These conditions appear to 

 be quite constant among the Teleosteans. Among a great number of 

 very diverse forms I have always found either an olfactory nerve in the 

 orbit or a long tractus extending directly from the brain case to the nasal 

 pit. 



In Hydrocyon^ already referred to, the bulbus lies in a special eleva- 

 tion in the orbito-sphenoid ; from it a nerve is given off that passes to 

 the olfactory membrane, being free in the orbital cavity ; and a long 

 tractus lying within the cavum cranii to the fore-brain, so that in this 

 case there is no exception to the general rule. 



A remarkable exception to this rule is found in all the Ganoids. In 

 these fishes a true olfactory nerve passes within the direct continuation 

 of the brain case, and consequently proves to be a condition that must 

 have arisen under circumstances to us nearly unknown and entirely 



