8 DISCOPHOKiE. Part III. 



characters thus far assigned to the Rhizostomeae, the Cepheos, and the Cassiopea?, 

 they differ most strikingly in their form, and especially in the form of their oral 

 appendages; that similar differences exist in the form of the Aurelia3, the Pelagian, 

 and the Cyaneoe ; and that the Charybdeoj are still further removed from these two 

 groups by their peculiar form, — the question at once arises, What are the characters 

 which bind the Rhizostomeaj, the Cepheie, and the CassiopeiB so closely that Esch- 

 scholtz should have united them as one natural group, even though he himself 

 never had an opportunity of examining any of their number ? and what are the 

 characters which justified Tilesius in dividing them into three families ? On the 

 contrary. What are the characters which led Eschscholtz to unite the Aureliaj, the 

 Pelagia), and the Cyaneae into one group, which is natural, even though the attempts 

 of recent writers to subdivide them into several families be equally justifiable ? 

 and what, finally, are the reasons which could satisfy Gegenbaur that the iEginidaj 

 are the most aberrant type among the Craspedota, though among themselves they 

 are very closely linked together ? 



I believe that these questions are not difficult to answer, if w^e apply to their 

 solution the tests which I have proposed in analyzing the different categories of 

 structure upon Avhich different kinds of natural divisions may be founded in the 

 animal kingdom. The Ehizostomeae, the Cepheee, and the Cassiopese may be dis- 

 tinguished as natural families because their form is different : they may be united 

 into one natural group because they agree in certain complications of their structure, 

 by which they at the same time differ from the Aurelia?, the Pelagite, and the 

 Cyanea3. These again agree with one another in some other complications of struc- 

 ture as much as they differ from one another in their form ; and this is also true 

 of the Charybdece and ^ginidse, which, as I shall show hereafter, ought to be united 

 into one and the same group, on account of the peculiar complication of their 

 structure, though they also constitute distinct families, characterized by their form. 

 We have thus among Discophora? proper, two categories of characters thus far not 

 sufficiently distinguished, which, when properly analyzed, lead to the recognition of 

 a greater number of natural families than are generally admitted among these Aca- 

 lephs, and at the same time point out the manner in which these families may 

 be combined into higher groups. But what are these higher groups ? Can they 

 be orders ? 



We have already seen that the class of Acalephs contains only three natural 

 orders, — the Ctenophora^, the Discophorre proper, and the Hydroidas, — characterized 

 by the complication of their structure, and occupying respectively the rank in which 

 they are here enumerated, the Ctenophoras being the highest and the Hydroidse 

 the lowest. If, then, there are among the Discophora3 natural groups of a higher 

 rank than families, and yet not entitled to be considered as distinct orders, they 



