INTRODUCTION. 



medusae and to exclude those which produce sessile gonophores. Nevertheless 

 this should be clearly understood and must be accepted as an artificial limitation 

 of the work. 



I have thus attempted to describe only such hydroids as are known to produce 

 medusae, and have endeavored to bring the systematic arrangement of the medusa 

 more nearly into accord with that of the hydroids. A strictly natural system includ- 

 ing both hydroids and medusae can not be constructed, for many of the hydroids 

 remain undetermined. Moreover, dissimilar hydroids (Svm-tjrv>if, StanriJinim 

 may give rise to similar medusae (Sarsia); or the reverse may be the case, as in the 

 medusae of Bongamvillia and Nemopsis, or that of the two sorts of medusa- (Sarsia 

 and Cladoncrna] which may arise from hydroids of Stauridia. 



These and many other cases of a similar nature interpose a barrier to our 

 attempts to invent a natural system which includes all hydroids and medusa" within 

 its embrace. At present, I believe, we must content ourselves with a compromise 

 between a natural and an artificial arrangement, confiding in the belief that as 

 more and more of the hydroids are discovered it will become correspondingly 

 more possible to arrange the medusae in a natural system. After consultation with 

 Prof. C. C. Nutting we have mutually decided that the promulgation of such ;i 

 system is at present inadvisable. Such a system has, indeed, been proposed by 

 von Lendenfeld, 1884 (Zool. Anzeiger, Bd. 7), but has gained no acceptance. 



Much confusion has been introduced through the habit, in vogue among 

 marine expeditions, of sending all of the medusa" to one specialist and the hydroids 

 to another. Thus the sessile and the reproductive stages of the same animals are 

 worked upon independently from different view-points by different men. 



I am inclined to regard the Trachymedusae and Narcomedusa" as being 

 transformed actinutae, for they commonly develop through an actinula larva in 

 which the bell grows out as a collar-like, or intertentacular lappeted expansion 

 from the sides of the body after the tentacles have appeared, and the tentacles of 

 the actinula become those of the medusa. The medusa of the Anthomedusas and 

 Leptomedusae is formed upon a different plan, for the tentacles grow outward from 

 the bell-margin after the bell has developed. I believe, therefore, that the bell of 

 the Trachymedusae and Narcomedusae is not homologous with that of the Antho- 

 medusa and Leptomedusae. It is evident that the entodermal otoliths of the 

 Trachymedusa" and Narcomedusae are not homologous with the ectodermal otoliths 

 of Leptomedusae. Budding and alternation of generations occur in both classes 

 of veiled medusae. 



I believe that the medusa-shape has been acquired independently in the 

 Trachyhna and Leptolma forms of veiled medusae. 



The colored plates contained in this volume consist of drawings, from life, of 

 medusae of the Atlantic coast of the United States. The text-figures, on the other 

 hand, are chiefly outline tracings from the illustrations of many authors; and are 

 presented in order to spare the reader the trouble of consulting numerous scattered 

 works of reference. These outline copies of previously-published drawings of 

 medusae were carefully traced from the originals by Mr. Carl Kellner, artist of the 

 Tortugas Marine Laboratory of the Carnegie Institution. 



Other outline figures are from life, and the majority of these were drawn by 

 the author while studying at Mousehole, Cornwall, England, and at the Naples 

 Zoological Station during the autumn and winter of 1907 and 1908. 



This book aims to be something more than an old-fashioned systematic treatise, 

 for it attempts to record, if not to review, all works upon the embryology, cytology, 

 oecology, physiology, etc., of all forms coming within the scope of the text. 



