126 PEOP. p. MARTiisr Duncan's eevision or the 



The genus Latimaandra is a very large one, and follows the 

 rule which is noticed under such circumstances ; it is ill defined 

 from some others. During the growth of vigorous individuals 

 remarkable changes occur in the arrangement and relation of 

 the calices. The union of the elongate calices by their walls 

 with their neighbours is the completion of the generic definition ; 

 but it sometimes does not occur everywhere in the same large 

 corallum or colony, and here and there the walls are not 

 united, there being a distinct crevice between them. "Were a 

 fossil specimen to be broken and part of it to retain the 

 true Latimeeandroid structure and the rest to show the incom- 

 plete union, the latter would be considered generically distinct 

 from the former by M. de Eromentel, and would be called 

 Chorisastrcea. 



Now A. E. von Eeuss described and figured, in his description 

 of the Possil Anthozoa of the strata of Castelgomberto*, beautiful 

 specimens of Latimceandra circumscripta, L. marcJielloides, and 

 L. dcedalcBa, in which the true character is present ; and also 

 equally well-preserved types of L. discrepans and L. dimorpha, 

 in which the walls are separate (plates 5-8). No one can doubt 

 that all these forms belong to one genus, and that there is no 

 necessity for the experiment of introducing Chorisastrcsa to 

 separate the last from the first. 



The same author carried the matter further, for he describes 

 {op. cit. p. 20, pi. V. figs. 2-3) a type in which there are calices 

 of the true Latimseandroid type, others separate and Choris- 

 astrsean, and many others which are circular in outline and free 

 all round to a considerable depth, looking like a combination of 

 Thecosmilia, Latimceandra, and Ohorisastrcea. The specimens are 

 large, and their habit of irregular growth and irrregular calicular 

 junction is most suggestive in a classificatory sense, because 

 there is a constant springing up of independent buds which 

 sooner or later become serial in their calices , and at last unite by 

 their walls. The colony is never entirely Latimseandran, Choris- 

 astrsean, or Thecosmilian. 



It is the independence of the circular calices, which is so unlike 

 anything Latimseandran, that characterizes amongst other things 

 Seterogyra, Eeuss. The type S. lobata, Eeuss, cannot possibly 

 come within the genera already mentioned. 



* "Pal. Stud, iiber die alt. tertiarsch. der Alpen," Denkschr. d. Kais. Akad, 

 Wiss. Wien, 1867-68, p. 21 et seq. 



