OBELIA. 14-7 



ferred by Peron and Lesuenr at a very early date on a 

 supposed Medusa, which we now know to be the repro- 

 ductive element of a Campanularian zoophyte. Agassiz 

 has adopted it for one section of the species, producing 

 gonozooids with a shallow, disk-like umbrella; and All- 

 man has extended it to the whole group. In this, I have 

 no doubt, he is right. The character employed by 

 Agasssiz to distinguish his genera Eucope and Obelia 

 (the number of arms which the sexual zooid possesses at 

 the time of liberation) is barely of specific value. 



The younger Agassiz insists upon the presence or ab- 

 sence of sporosacs at the time of liberation as an important 

 character, and goes so far as to separate the 0. yemculata 

 of Wright from that of Gosse because they differ in this 

 respect. To the former, which he identifies with the 

 Eucope diaphana of his father's great work, he assigns 

 the name of alternata. I feel convinced that this is a 

 mistaken view ; and unfortunately it is one which intro- 

 duces confusion and difficulty into the science by enlarging 

 the list of synonyms. The fact is, that the gonozooids 

 contained in the same capsule leave it in very various 

 stages of development ; this is proved by the differences 

 in size and the condition of the marginal tentacles at the 

 time of liberation*. And I have no doubt that the earlier 

 or later appearance of the ovaries is to be explained in the 

 same way : it is a variation in the degree of development, 

 and not a specific difference. On the free zooids of Clytia 

 Johnstoni Wright and myself have found ovaries with ova 

 directly after their escape ; Allman, on the contrary, met 

 with none on his specimens. A. Agassiz found them in 



* Wright also mentions another diversity amongst the gonozooids of 0. 

 geniculata. In some the ovaries were close to the manubrium, in others 

 midway between the base of it and the marginal canal ("Observat. on Brif. 

 Zoophytes," Ed. N. P. Journ. for .Tnn. 1*.">9). 



