REPORT ON CORALS — HYDROCORALLIN^E. 99 



pores sporadic, with tabulae and without styles, and with two kinds of zooids, with 

 knob-bearing tentacles ; with a tendency also in the dactylozooids to form ring-like 

 groups around a gastrozooid. This form may be termed " Archihydrocorallina." 



Archihydrocorallina was probably derived from a form in which all the zooids com- 

 posing the stock were provided with mouths and generative organs. In such stocks 

 further development may be conceived of as having arisen by either of two processes. All 

 the zooids may have become gradually modified, so that each performed only one function, 

 and thus had certain of its structures aborted to fit it for this special end. If such be the 

 history of the development of the Hydrocorallinse, then the gastrozooids, dactylozooids, 

 and generative zooids are to be looked on, as they have been regarded throughout the 

 present memoir, as zooids which have become more or less rudimentary by disease. Or, on 

 the other hand, the view may be taken that the gastrozooids alone represent the original 

 zooids of the ancestral stocks. They remain, having lost their generative organs, and. 

 to a greater or less extent, their prehensile ones, because additional zooids have been 

 formed by budding in order to provide for the wants of the colony in these particulars. 

 On this view the generative zooids and dactylozooids were originally budded out in the 

 condition in which they now exist, or in one not so complete as it is at present, nor so 

 perfectly adapted to their present functions. On this view they have lost no structure 

 by disuse, but have rather advanced in complexity with development, but only in their 

 own special direction. 



The former view of the antecedent history of the sub-order Hydrocorallinre seems to 

 me to be most worthy of acceptation, because the presence of several structures which 

 occur as rudiments in connection with the dactylozooids and generative zooids, but which 

 are fully developed in connection with the gastrozooids, seems to bear out this conclu- 

 sion. As examples, may be cited the tentacles of the dactylozooids of the Milleporidae and 

 the styles of the dactylozooids of certain Stylastemke. The Stylasteridse in the com- 

 plexity of their compound stocks form an interesting parallel to the Siphonophora. In 

 the Siphonophora the several components of the compound organisms are by the best 

 authorities regarded not as individual degenerate zooids, but as buds which tend to 

 assume more and more the form of individuals. The diverse elements composing the 

 organism in the case of the Siphonophora may seem closely paralleled by those of which 

 a Stylastericl is made up, and yet the past history of the two organisms may be very 

 different. In the one case, an ancestral already compound organism may have gradually 

 modified its similar zooids to subserve division of labour ; whilst, in the other, a simple 

 ancestor may have gradually developed a similar compound organism by throwing out 

 buds of various forms which have come more or less to approach itself in complexity. 



From Archihydrocorallina, Archistylaster was developed with a branching ccenosteum ; 

 with a strong tendency to assume a flabellate form, and to develop its pores only on one 

 face of the flabellum, and at the sides only of the branches ; with its pores sporadic and 



