Reviews — Development of Monticuliporoids. 33 



palaeontologists nowadays place tliem, usually with some hesitation, 

 among the Polyzoa. Eastman in his translation of Zittel's Grundzuge 

 der Paldontologie takes the original course of placing them under 

 both Corals and Polyzoa. Still more startled were palaeontologists 

 when a letter written by Kirkpatrick to Nature appeared,^ stating 

 that Monticuliporoids were allied to Merlin, a recent siliceous Sponge. 

 "The discovery," says Kirkpatrick, " of the solution of the problem 

 of Merlia is destined to prove of profound importance to palaeontologists. 

 For I have now convincing proofs . . . that nuinei-ous Palaeozoic 

 fossils coming under the old-fashioned term ' Monticulipora ' are 

 essentiall}^ the same nature as Merlia, and that they are supplementary 

 calcareous skeletons of siliceous sponges. Merlia seems to be a solitarj' 

 survivor of the Monticulipora type from Palaeozoic times . . ." 



Now it was obvious that, if Merlia were a siliceous Sponge and the 

 Monticuliporoids allied to it, the post-embryonic development would 

 be similar in both cases, whereas if Monticuliporoids were Polyzoa 

 they should present post-embryonic stages characteristic of that 

 phylum. What was required, then, was the determination of the early 

 stages of Monticuliporoids. This was meanwhile being done, quite 

 independently of anj' thought of Merlia, by Cumings in America, 

 with the result that, in the same month as Kirkpatrick's letter in 

 Nature appeared, there was published an account by Cumings of the 

 early stages of certain Monticuliporoids, showing them to agree closely 

 in this direction with Cyclostome Polyzoa. 



The genera in which Cumings has shownPolyzoan growth-stages from 

 the time when the larva first became fixed are three — Prasopora, 

 Phylloporina, and Callopora ; but in four other genera he has found 

 very early, though not the earliest, stages, which leave no doubt that 

 these four genera are of the same nature as the first three. It is 

 unfortunate that Cumings makes no mention of the earliest stages of 

 Monticulipora mammulata, the genotype of Monticulipora chosen by 

 Nicholson ^ from the four genosyntypes presented by d'Orbigny.^ 

 For it is possible, though unlikely, that included among the Monticuli- 

 poroids are forms not related to the Polyzoa; and it is the genotype 

 of Monticulipora that Kirkpatrick claims as related to Merlia. 

 Now, should this prove to be anything but a Polyzoan, the name 

 Monticuliporoid, used, apparently, by Cumings as synonymous with 

 Trepostome, will have to be abandoned to whatever group includes 

 Monticulipora mammulata. 



If, as is likely, Monticulipora mammulata\?,^n:o\ediio\)e?,jm^\\\\Qt\c 

 with the genera that Cumings has conclusively shown to be 

 Trepostomes, it presents in its skeleton a very pretty example of 

 horaoeomorphy with Merlia. 



It is impoi'tant to appreciate the nature of Cumings' evidence. 

 Four Polyzoan orders are well known as fossils, namely, Cyclostomata, 

 Tiepostoraata (including Monticuliporoids), Cryptostoraata, and 



1 Nature, vol. Ixxxix, p. 502, 1912. 



2 Nicholson, On the Structure and Affinities of the genus Monticulipora ajid 

 its suh-genera, 1881, p. 1. 



^ D'Orbigny, Frodrome de PaUontologie stratigraphigue universelle, vol. i, 

 p. 25, 1850. 



DECADE V. — VOL. X.— NO. I. 3 



