42 MORPHOLOGY. 



termination in the circular canal. The tentacles, which in many medusae spring from the 

 intervening spaces upon the margin of the umbrella, and arc therefore not directly continuous 

 with the radiating canals, make their appearance probably in all cases later than the others, and 

 arc frequently less developed. These inter-radial tentacles must be placed in the same category 

 with the lithocysts as simple marginal appendages, to be carefully distinguished from the primary 

 tentacles, and, like the lithocysts, have no representative in the hydranth. 



It cannot be urged, as an argument against this view, that the circular canal of the medusa is 

 not represented in the hydranth ; for the absence of a developed umbrella in the hydranth neces- 

 sarily brings with it the absence of this canal ; and it is for the same reason that velum, lithocysts, 

 and secondary tentacles, are also absent. Neither can it be said that those cases in which 

 the tentacles of the hydranth are not arranged in a single verticil, but are repeated regularly or 

 irregularly in different planes upon the body, are inconsistent with the homological relations here 

 insisted on ; for such cases can be regarded only as special modifications of the more typical plan 

 which has directly suggested our comparison. 



Huxley, believing the difference in structure and development between the locomotive disc 

 of the gymnophthalmic and that of the steganophthalmic medusae to be so great as to place them 

 in different categories, would confine the term " umbrella" to the disc of the stef/a/iophfJMlmafa, 

 and would designate that of the (/i/mjiophfhalwafa by the terms " nectocalyx" and " gonocalyx." I 

 was at first disj)osed to adopt the same view ; but an investigation of the mode in which this part 

 makes its appearance in the gymnophthalmic forms has convinced me that the development is 

 essentially the same in both cases, and that, notwithstanding some marked structural differences, 

 there is sufficient unity between the two to render it more convenient to speak of them under the 

 same term as strictly homologous organs. In both cases they are formed by an outgrowth of the 

 walls of the polypoid manubrium, and the fact that the steganophthalmic medusa is produced by 

 successive transverse divisions of a " scyphostoma," while the gynmophthalmic medusa is formed 

 as a lateral bud from a hydrosome, is no valid argument against this approximation ; for 

 every segment of the " scyphostoma" is strictly comparable to the bud of the hydroid, and 

 developes its umbrella by an outgrowth from its sides in quite the same way. 



A very instructive example, which strikingly bears out the comparison I have here attempted 

 to make between the hydranth and the medusa, is afforded by the remarkable locomotive zooid 

 which forms the gonophore of Dicomjne. This little zooid (woodcut, fig. G) is essentially a free 

 medusa, reduced to the condition of an ova-bearing or spermatozoa-bearing manubrium, from whose 

 base two free tentacnla in form and relations like those of a hydranth are developed. Now, there 

 is here no umbrella, locomotion being affected by the action of cilia ; but it is evident that we have 

 only to imagine the ectoderm of the manubrium projected as a disc, in the way already supposed, 

 in the horizontal plane passing through the base of the two tentacles so as to include the basal 

 portion of these tentacles in its thickness, in order to have an umbrella with two radiating canals 

 added to the manubrium. 



But development, as we shall afterwards sec, entirely coincides with anatomy in pointing to 

 the same conclusion ; and it is only necessary to trace the formation of the umbrella and radiating 

 canals in the budding medusa, in order to become convinced that their origin is essentially that 

 here insisted on) ; while the interesting observations of Johannes Midler on the development 

 oi yEginopsis, and of M'Crady on that of Cimina — observations which will be specially referred to 

 below — show tliat in these genera the umbrella grows out as a horizontal disc from the walls 



