Ix 



INTRODUCTION. 



present no structural peculiarities that would entitle them 

 to stand alone, and are rightly merged in the Order, which 

 includes so many kindred zooidal forms. Agassiz and 

 Fritz Miiller have taken this view"; and Cams, in his 

 admirable classification of the Ccelenterata, has referred 

 them to the Hydroida, though he has placed them in a 

 distinct group (Haplomorpha) , apart from the forms into 

 whose life-history the two elements enter. Huxley pro- 

 poses a separate Order for the naked- eyed medusas that 

 are developed directly from the eggs of similar organisms ; 

 but the absence of the fixed-polypite stage can hardly be 

 accounted more than a generic character when it is re- 

 membered that the Lizzia observed by Claparede, the eggs 

 of which produce medusas,, is identical in structure with 

 the sexual zooid of the Campanularian LepioscypTms (All- 

 man). I can see no reason whatever for detaching the 

 medusan forms developed directly from the ovum, and not 

 as buds on a fixed stock, from the Hydroida, either as a 

 separate order, or even as a secondary section. To the 

 latter they are bound by the closest structural affinities ; 

 and instead of dismembering the Hydroid group on the 

 ground of this difference in the mode of development, it is 

 surely more philosophical to enlarge our conception of its 

 range. 



I have therefore rejected Carus's subgroups Haplo- 

 morpha and Diplomorpha, and have preserved the simple 

 unity of the order Hydroida. The present work, however, 

 embraces only the medusan forms that have been traced 

 to a fixed Hydroid stock. 



Another result to which we have been brought by our 

 increased knowledge of Ccelentcrate structure is the recog- 

 nition of the close affinity subsisting between the Siphono- 



