86 MANUAL OF ZOOLOGY. 



calyx depend marginal tentacles, which are usually hollow 

 processes, composed of both ectoderm and endoderm, and in 

 immediate connection with the canal system. Also round the 

 circumference of the nectoealyx are disposed certain 'mar- 

 ginal bodies,' of which two kinds may be distinguished. Of 

 these the first are termed 'vesicles,' and consist of rounded 

 sacs lined by epithelium, and containing one or more solid, 

 motionless concretions apparently of carbonate of lime 

 immersed in a transparent fluid. The second class of marginal 

 bodies, variously termed ' pigment spots,' ' eye-specks,' or 

 'ocelli,' consists of little aggregations of pigment enclosed in 

 distinct cavities. The ' vesicles ' are probably rudimentary 

 organs of hearing, and possibly the eye-specks are a rudi- 

 mentary form of visual apparatus. The oral margin of the 

 polypite may be simple, or it may be produced into lobes, 

 which are most frequently four in number. The essential 

 elements of generation are produced in simple expansions 

 either of the wall of the manubrium or of the radiating necto- 

 calycine canals. 



From the above description it will be evident that the 

 Medusa is in all essential respects identical in structure with 

 the free- swimming generative bud or gonophore of many of 

 the fixed and oceanic Hydrozoa. Indeed, a great many forms 

 which were previously included in the Medusidce have now 

 been proved to be really of this nature, and it may fairly be 

 doubted if this will not ultimately be found to apply to all. 

 As to the value, however, of the order Medusidce, the present 

 state of our knowledge is well expressed by the following con- 

 clusions which have been drawn up by Professor Greene : 



' 1. That several of the organisms formerly described as 

 Medusidce are the free gonophores of other orders of Hydrozoa. 



' 2. That the homology of these free gonophores with those 

 simple expansions of the body- wall which in Hydra and some 

 other genera are known to be reproductive organs by their 

 contents alone, is proved alike by the existence of numerous 

 transitional forms and by an appeal to the phenomena of their 

 development. 



' 3. That many of the so-called Medusidce may, from analogy, 

 be regarded as, in like manner, medusiform gonophores. 



' 4. But that there may exist, nevertheless, a group of 

 Medusid forms, which may give rise by true reproduction to 

 organisms directly resembling their parents, and therefore 

 worthy of being placed in a separate order under the name 

 Medumdce.' 



The same authority concludes by remarking that to the order, 

 as above defined, ' may be referred provisionally that large 



