1883.] MADREPORARIAN GENUS PHYMASTR^A. 407 



sequently the diagnosis of the genus was given in their work entitled 

 'Ilecherches sur la Structure etla Classification des Polypiers recents 

 et fossiles ;' and finally in their great work, the ' Ilistoire naturelle 

 des Coralliaires,' tome ii. p. 499 (1857). 



The generic diagnosis of 1857 does not correspond in a very im- 

 portant point with that published in 1848; and there is no doubt 

 that the last diagnosis is erroneous. The great French zoophytolo- 

 gists described two species of the genus PhiimastrcBa in their last- 

 mentioned work, having noticed them fully in their previous one. 

 Unfortunately the geographical positions of the two species are not 

 known. Probably they are from the Eastern seas. 



II. The Generic Z)ea^MO«es oy Phymastrsea' given in 1848 awe? 1857. 



The generic diagnosis given by MM. Milne-Edwards and Jules 

 Haime in 1848 is as follows : — 



" Corallum in a convex or plane mass. Corallites prismatic and en- 

 veloped from the base to the summit by a thin epitheca without a 

 trace of costee. Corallites close, not joined by thin walls but, at 

 certain distances, by large wart-like projections, so that there are 

 considerable spaces between the walls of contiguous corallites. The 

 gemmation is extracalicular and subapical. The walls are thick ; 

 the calices are subpolygonal, and their margins are free ; the colu- 

 mella is spongy in texture and well developed ; and the septa are well 

 developed, slightly exsert, and strongly toothed, especially near the 

 columella." 



The generic diagnosis published in 1857 contains the same cha- 

 racters, but there is an alteration regarding the method of increase 

 of the corallum in the nature of the gemmation. This is stated to 

 be calicular and submarginal. 



Fortunately the drawings and descriptions of the species of Phy- 

 masti-cea which were published along with the first diagnosis of the 

 genus will satisfy any student of the Madreporaria that the gemma- 

 tion is extracalicular, and from the wall beyond the edge of the 

 calices. The second diagnosis is therefore incorrect ; and this opinion 

 is maintained after the examination of the third and hitherto un- 

 published species. 



In noticing the remarkable method of the junction of the corallites 

 of Phymastrcea, MM. Milne-Edwards and Jules Haime state that 

 the genus has great affinities with others of the Astrseidse, and that 

 this method is characteristic. 



It is necessary to draw attention to the statement that the coral- 

 lites are invested with a thin epitheca and that there is no trace of 

 costse. It is evident, moreover, that MM. Milne-Edwards and Jules 

 Haime considered the junction-processes to be invested with epitheca, 

 but to consist of an almost compact structure. These processes 

 "se soudent fortement d la muraille d'un polypierite voisin." 



In the delineation of the species Phymastrcea valenciennesi, Ed. & 

 H., on pi. ix. figs. 3 & 3 ffl (Ann. des Sciences Nat. 3 serie, t. x., Zool.), 



' (j-vfia, a thing that grows upon the bod}'. 



