ME. H. M. BEENAED ON RECENT POEITID^. 137 



wLellier tlie internal theca of Porites has been developed out of 

 the exsert septal spines of a Favositid, or is a secondary modifi- 

 cation of that of the Madreporidse, with its ring of lamellated 

 septa. 



Recent work with both Goniopora and Porites has led me to 

 the conclusion that this latter view is the correct one. The 

 septa in both these two genera were once purely lamellate. Proof 

 of this can be seen in the fact that, both in Porites and in 

 Goniopora, every transition can be found between the forms with 

 almost purely lamellate septa and others with purely " trabecular" 

 septa. J'urther, in a great many forms, the costae round the 

 growing edges run out as vertical lamellae to the rim of the 

 epitheca, that is, just where primitive conditions might be 

 expected. 



This conclusion is not only of permanent importance for the 

 solution of the problem we have in hand, but it deserves the 

 special attention ot all students of Stony Corals, because it relieves 

 them for ever of the " trabecula " as a unit of morphological value. 

 The " trabecula," which is in reality merely so much formative 

 tissue, was brought into the system by the theoretical scheme 

 of Madreporarian tectonics put forward by Milne-Edwards and 

 Hairae, a scheme wliich Miss Ogilvie * has recently endeavoured, 

 by considerable emendation and amplification, to place on 

 the surer basis of extended histological research. How great a 

 snare it has been I have already shown t in the case oiMontipora, 

 which, on account of its " trabeculae," was ranked by Milne- 

 Edwards and Haime among the Poritidae. The " trabeculae" of 

 Montipora, that is, if what I called the trabeculae of that genus 

 are what Milne-Edwards meant, turned out on examination to 

 be very different from those of Porites (see further on this point 

 below). Again, if the trabeculae in Porites and Goniopora, 

 in which genera of all others they appear to play the most 

 important part, can yet be shown to have no real morphological 

 value, their case finally breaks down. 



My own experience is as follows. I began work first with 

 Goniopora, its larger calieles admitting of easier examination. 

 At the outset the " trabecula " was accepted as a morphological 



* Phil. Trans, vol. 187, 1896. 



t Ann. & Mag. Nat. Hist. xx. 1897. 



