CAMPANULARIID^. 139 



we have free medusiform zooids, and these exhibit different 

 types of structure. 



The generic groups in this family are founded for 

 the most part on the differences presented by the 

 reproductive system. In Lovenella alone are good dis- 

 tinctive characters supplied by the polypite and caly- 

 cle; in the remaining genera they must be sought al- 

 together in the portions of structure connected with the 

 propagation of the species. As these are not always 

 present, a practical difficulty will often encounter the 

 student ; he may be unable, in the first instance, to refer 

 his specimen to its genus, and must go through the 

 species of several allied genera in order to determine its 

 position. A similar difficulty meets the systematist, who 

 can only assign a provisional place to those forms the 

 reproductive phase of which has not been observed*. In 

 one or two cases I have ventured to locate such forms 

 conjecturally; but most of them are ranged for the time 

 under Campanularia, with which, so far as the trophosome 

 is concerned, they agree, and must be referred to their 

 proper place in the system as their reproductive history 

 is known. The number of such species is but small. 

 Much more numerous probably are the reproductive 

 zooids that have not yet been traced to the Hydroid 

 stock. 



The Campanulariida are very widely distributed over 

 the globe. Many of the British forms have been obtained 

 on the Atlantic coast of North America, as well as a con- 

 siderable number of species belonging to the genera Clytia 

 and ObeHa which have not been found in our seas. One 

 or two fine species of Campanularia are described by A. 

 Agassiz from the Pacific side one of them, of gigantic 

 size, ranging from San Francisco to Behring's Straits. 

 * These remarks apply to many other families. 



