8 MADEEPORAEIA. 



according to the way the septa were united to form thecse. (1) Athecalia, when the union was 

 iiscontinuous and brought about by scattered synapticulae resulting in a kind of porous theca ; 

 (2) Euthecalia, in which the septa were united by a continuous outer wall from the inner 

 surfaces of which they projected ; and (3) PseudothecaUa, in which the theca was continuous 

 but not a distinct structure, being due] to fused lateral thickenings of the septa. This new 

 principle of classification necessitated a recasting of the scheme sketched in 1887. This 

 affected the Poritidse, which had now to be classified as of the order Athecalia, sub-order 

 Madreporaceas, of equal value mth Thamnastrisacese and Fungiaceae. The Poritidse belong to 

 the Madreporacese and contain the genera Psammocora, Synarcea, Pontes. The genus 

 Goniopora is not mentioned. A new genealogical tree is given which again deduces Porites 

 with Astrceojjora from Actinarcea and Microsolena. The relationship of the genus to 

 Madrepora is close but left uncertain. With regard to this whole system, we may again 

 remark that the wall taken alone is just as unable to supply us with a classification as are the 

 septa relied upon by Milaschewitch and Pratz (see above). 



In 1896, Miss Ogilvie* again took up the Poutrelle (trabecula) of Milne-Edwards and 

 Haime, and extending the work of Pratz sought to make the fine structure of the septa the 

 fundamental basis of classification. On these lines, however, she found it impossible to 

 decide whether the Poritidse were or were not related to the Madreporidae, or were an 

 independent development. On page 340 Miss Ogilvie suggested dividing the Poritidse into 

 Poritinse and Spongiomorphinse Freeh. Among the ancestors of the Poritidse we find 

 suggested the following genera : Litharcea and Dictyarcea (Tert.), Astrwomorpha and Heptastylis 

 (Trias.), Protarcea, Stylarcea, Coccoseris, and TJiecia (Pal.). (See list on the opposite page.) 



In 1899, t the present writer introduced a new character for the classification of the 

 Stony Corals, viz. the primitive epithecal cup. The close relationship of the Poritidse and 

 Madreporidse with the Eupsammiidse (Perforata M.-E. & H.) was entirely confirmed, for 

 in all of them the epithecal cup is early flattened out, and the subsequent skeleton is built 

 up solely of the septa with their junctions. The family Poritidse was limited to two 

 genera, Porites and Goniopora, and was derived from Madreporids which budded very 

 early, i.e. while the skeleton was still an immature basal structure. This view is in the main 

 adhered to in this volume, but the proposed relationship of the genus Goniopora to Porites has 

 been reversed, owing to the discovery of the septal formula of Goniopora and the fact that 

 it passes into the septal formula of Porites by a process of reduction. 



The following is a list of genera which have been claimed as Poritidse, or suggested as 

 possible ancestral forms, and here either excluded altogether, or merged with one of the two 

 genera Goniopora or Porites. 



1. Adinarcea D'Orb. (= Agaricia granulaia, Goldfuss, Petrel, pi. xxxviii. fig. 4), suppressed by 



Duncan as = Microsolena, in his 'Revision of the Coral System of Milne-Edwards and 

 Haime,' Journ. Linn. Soc, xviii, (1884) p. 168. 



2. Alveopm-a Quoy and Gaimard, Voy. de 1' Astrolabe, iv, (1833) p. 238. A Favositid, vide Journ. 

 Linn. Soc, xxvi. p. 49-5. 



* Phil. Trans., clxxxvii. p. 326. f Journ. Linn. Soc, xxvii. pp. 128, 487. 



