1848.] LONSDALE ON FOSSIL ZOOPHYTES. 79 



1. Orhicella, in which the *' cells" (stellated cavities) do not sub- 

 divide, or rarely, and appear to be tubular vt^ith interstices. 



2. Siderina, in which the *' cells" do not subdivide, nor appear to 

 be tubular. 



3. Fissicella, in which " the cells subdivide by growth and bud- 

 ding." 



In this arrangement therefore corals are generically re-united, 

 though the productive processes differ, being non-iissiparous in Side- 

 rina, rarely fissiparous in OrhiceUay and essentially so in Fissicella. 

 Lastly, Mr. Gray of the British Museum, in a paper published, Fe- 

 bruary 1847, in the Annals of Natural History, '*on an arrangement 

 of Stony Corals," doubts where the Astrcea of Lamarck should be 

 placed ; but introduces the genus provisionally (?) into the family 

 Agariciadce, one division of a great group or order characterized by 

 "the animal growing by spontaneous division." {op. cit. p. 128.) 



This brief summary of opinions entertained by five systematic 

 writers, whose resources and means for forming a right judgement 

 have been most extensive, leads to the inference, that the essential 

 characters of Astrcea, even as respects existing species, remain to be 

 fixed. The abstract shows, nevertheless, that the latest authorities 

 coincide in regarding the fissiparous process as an element of the 

 genus ; and it is believed that all the corals hitherto considered as 

 Astrcece, in which it does not occur, or so rarely as to be an excep- 

 tion, possibly an accidental condition, should be removed to another 

 family or order. 



But the palaeontologist must not consider a fissiparous increase in 

 a circumscribed star as sufficient to prove that the fossil in which it 

 is detected is an Astrcea. Among existing globular corals with per- 

 fectly bounded cavities, the operation is not effected uniformly ; and 

 the variations in connection with other structural peculiarities may 

 be found valuable in proposing generic or subgeneric groups. In a 

 recent coral, believed to be the Astrcea {Favastrcea^ magnijica of 

 M. de Blainville*, the divisional process commences by the develop- 

 ment of a reticulation among the lamellae, or the forming on one side 

 of the cavity of a new centre, equivalent in position with an additional 

 digestive apparatus. It is not a partition from a previous reticulated 

 centre, but a distinct structure separated by a clear, simply lamellated 

 interval. Subsequently a divisional barrier is constructed, towards 

 which the young digestive organs evidently supply a modicum of 

 materials. Ast. magnijica belongs to Ehrenberg's ixw^ Astrcece \ and 

 the stars expand as they grow upwards, so as to leave no interspace; 

 and it is impossible to conceive how, under such conditions, a new 

 cavity could be formed, except within the area of one previously 

 existing. Astrcea porcata of Lamarck {op. cit. p. 406, No. 7)t 

 affords an example of a different mode of effecting a subdivision. In 



* Man. d'Actinol. p. 374, Group. I. pi. 54. fig. 3 ; consult also Mr. Dana, op. cit. 

 p. 231. 



t Ast. {Meandriniforma) id., De Blainville, op. cit. p. 3G7. Favia of Ehren- 

 berg, Beitrjige, p. 94. Consult Esper's Pflanzenthiere, Mad. pi. 71 ; also Dana, 

 Zooph. p. 226. 



