Mode of Life among Medusce. 367 



loss of tentacles, the polymorpliism, and the increase in pro- 

 minence of the sexual bodies, are exactly what we should 

 expect to find a priori if a degradation had taken place in its 

 structure. 



There is one other point to which I wish to call atten- 

 tion before closing my communication. The existence of a 

 polymorphism, such as we find in Hydrichthys^ is exceptional 

 among fixed Hydroids of the Tubularian group. Something 

 similar exists in Hydr actinia and Perigonimus and one or 

 two other genera ; but this kind of polymorphism is not com- 

 mon among fixed Hydromedusas. A similar polymorphism 

 exists, however, in Velella, a floating Hydroid well known 

 to all naturalists. In Velella we have the basal plate with 

 anastomosing tubes of Hydrichlhys modified into a compli- 

 cated float. The gonosomes are the same in both genera, 

 the filiform bodies of HydrichtJiys are represented by the 

 single central polyp, so-called, in Velella. The Meduste of 

 the two closely resemble each other. There are only two 

 kinds of individuals in both genera. 



Strangely enough, after I had reasoned out this likeness 

 between Velella and Hydricldhys on morphological grounds, 

 my memory went back to a strange story I had once heard 

 from an Italian fisherman of the orighi of Velella from the 

 common mackerel. This story or a similar one long ago 

 found its way into the books. 



According to Marcel de Serres, the Mediterranean fisher- 

 men suppose that Velella originates as a bud from the head 

 of the mackerel ; and Pagenstecher goes on to explain this 

 error, after quoting its source, from the fact that young 

 Velellce are often found in the nets with the fishes, and it is 

 easy to suppose, as their colour is similar, that one budded 

 from the other. While we accept without question this ex- 

 planation and the want of foundation of the fishermen's yarns, 

 it is a strange coincidence that a possible relative of Velella 

 should be found attached to the body of a fish. It is well 

 for us to inquire, in the light of ])hylogeny, whether Velella^ 

 if it has not itself originated from Hydroids on the fish bj 

 budding, has not been directly derived from one which is so 

 intimately related to Hydricldhys^ which is attached to the 

 body of a fish, that an unskilled observer might be easily 

 deceived. 



Hydrichthys is, in point of fact, the nearest known ally of 

 Velella among fixed Hydroids, and their morphological like- 

 nesses have already been pointed out. It would be prema- 

 ture to suppose, however, that Velella has derived its peculiar 

 anatomy from its descent from a form like the parasitic 



