128 PEOP. p. M. DUNCAl^ OliT THE STEUCTUEE AND 



their Hist. Nat. des Corall. vol. ii. p. 543 and 544, they finally 

 elaborated the genus Latimceandra. 



They did not see the necessity for eliminating the species the walls 

 of which are not perfectly united up to the margin of the series of 

 calicos, and therefore they placed in the genus such forms as L. Ber- 

 trandi, J. Haime, which has incompletely united series, L. Michelotti, 

 J. Haime, which has the series sometimes closely united and some- 

 times has them disunited, and L. Flemingi, Ed. & H., which has a 

 solid corallum with union of the valleys. 



Stoliczka's opinion is worth recording. Eeferring to Latimceandra 

 (" Cretaceous Corals of S. India," p. 37, Pal. Indica, 1873), he says : — 

 " There can be two sections distinguished, the one called Chorisas- 

 trcea by De Fromentel, in which the calices are separated from each 

 other by more or less distinct depressions, and the second, or Lati- 

 mcBandra, in which the series of calices are separated by united 

 ridges, over which the costse pass without interruption. These two 

 sections are so intimately connected with each other that authorities 

 on the subject, like Milne-Edwards or B,euss, do not consider a 

 separation into two genera practicable." 



To add to the confusion produced by the adaptation of the so-called 

 genus Chorisastrcea to serial Corals as well as to Latimceandrce which 

 increase by calicular budding, an attempt was made to include the 

 well-known species of Thecosmilia, which at first increase by gem- 

 mation and then by fissiparity. De Eromentel's generic diagnosis 

 did not permit of this extension, which has led to error. Eor 

 the drawing of C. gregaria, M'Coy, sp. (E. Tomes, op. cit. pi. xviii. 

 fig. 3), although it is figured to indicate the absence of union of the 

 walls of the specimen, the species being said to belong to Chorisastrcea, 

 shows the corallites in perfect lateral union. 



Finally the diagnosis of De Eromentel was altered by Mr. Tomes, 

 and it will be noticed in the ' Essay on the Lower-Oolite Madrepo- 

 raria,' p. 425, that Chorisastrcea has corallites " not divided by walls, 

 gemmation is basal, the corallum lobular, the lobes springing from 

 the base of the parent coraUite." 



These different diagnoses for one genus indicate that it has been 

 attempted to be founded on the methods of growth of structures 

 which are common to diverse genera and even families. 



Following Milne-Edwards and Eeuss, I restore the species in- 

 cluded in the genus Chorisastrcea to their former position in the 

 geuus Thecosmilia, and persist in declining to introduce into the 

 classification a genus which has three different diagnoses. 



At the close of the essay now in course of consideration its author 

 changes the name of Thecosmilia obtusa, d'Orb., to Chorisastrcea ob- 

 tusa. The species was illustrated in my ' Monograph,' Pal. Soc. 

 pt. iii. 1872, pi. i. figs. 1-4, and it will be perceived that the 

 increase is by fissiparity, and that there is no deep groove between 

 the walls of the primary and secondary calices. There is nothing 

 Chorisastrsean about the form, according to the generic definition of 

 M. de Fromentel, and I therefore restore the species to the genus 

 Thecosmilia. 



