, POEITIDiE. 7 



structure of the septa in Cyclolites, Tliamnastrwa, and a few other " Poritids " of Milaschewitch 

 (see last page). They are figured as apparently ideally trabecvdar in the sense originally 

 meant by Milne-Edwards and Hainie. 



Pratz did not, however, follow Milaschewitch in regarding all corals with such septa as 

 Poritidse. He would rather deduce the true Poritids from the Pseudastraeinse (= Thamnas- 

 traeinse and Cyclolitinse) which he regarded as Fungids. The nearest ally according to this 

 scheme is the genus Tliamnarcca Etall.* ; which, however, seems to me to be too insufficiently 

 diagnosed to be assigned any definite place in the coral system. With reference to Pratz's 

 work, we must again remark that the Stony Corals cannot be classified on one structure alone, 

 even though that structure is the septum. 



In 1884, Martin Duncan f published his revision of the Madreporarian system of Milne- 

 pjdwards and Haime. The Poritidse were divided into two alliances, Poritinoida and Monti- 

 poroida. The list includes all Milne-Edwards' recent genera (except Goscinarcea) together with 

 Sijnarma Verrill, which is placed next to Porites, and a new genus, Napopora Quelch. He did 

 not follow Klunzinger in reviving Stylarcea. Montipora (with the allied Anacropora Quelch) was 

 thus replaced among the Poritidse in face of the opinion of so many of the authors quoted. 



In describing the family he follows Dr. Klunzinger in stating that the corallites are joined 

 " by their trabeculate walls without intervening ccenenchyma " (see below, p. 19). 



In 1886, Mr. Quelch, in his report on the ' Challenger Eeef Corals,' as I think correctly 

 placed the Poritidce immediately after the Madreporidse. The same advance is not found in 

 the more detailed treatment of the family ; for Alveopora was retained, and Synarcea separated 

 from Porites by a new transitional genus Napofora (not here accepted, see list, page 9). A 

 second new Poritid genus Tichopora was established and placed after Goniopora and Ehodarcea. 

 This has also been rendered unnecessary as the range of specific variation absorbs it (see p. 16). 

 No description of the family was given by Mr. Quelch. 



In the year 1887, Dr. Ortmann f somewhat modified Pratz's conclusion. Dr. Ortmann's 

 Poritidse (= Madreporaria Perforata M.-E. & H.), include the sub-families Turbinarinse, Monti- 

 porince, Poritinse, Madreporinje, and Eupsamminte. Of the special origin and relationships of 

 the Poritiuffi he says little, but his larger Poritidse he would deduce through Adinarcea and 

 Microsolena from the family Thamnastrseidai, to which he gave far greater importance than 

 had Pratz, subordinating to it not only the Poritidae but also the Fungiidse. He expresses the 

 relationships of the great groups of the Hexacoralla in an interesting tree (1. c, p. 204). 



In 1888,§ the same writer again uses the word Poritidae in the wider sense above noticed, 

 but in the purely systematic parts of his paper he adopted the term Poritidae in its usually 

 limited sense. He may have intended the word " Poritinae." The included genera are the 

 same as those of Dr. Klunzinger, except that Synarcea is brought next to Porites. 



In the following year, 1889, the same writer || elaborated a new division of the Stony Corals 



* Lethiea Bruntrutana, 1864, p. 412. f Journ. Linn. Soc, xviii 



X Neues Jahrb. Mineralogie, 1887, pt. ii. p. 183. § Zool. Jahrb., iii. (Syst.) p. 143. 



II Zool. Jahrb., iv. (Syst.) p. 493. 



