4 i6 NATURAL SCIENCE. June, 



in a fraction of two or three years, to master the problems presented by 

 the literature alone. It is not surprising that the writer of this book has 

 repeated a few errors that seem to be the common property of text- 

 books. It will be well to correct these at some length, and with the 

 lucidity that must hitherto have been lacking in our attempts. 



First, we find this lamentable confusion between the dorso- 

 central and centrodorsal, as well as a misconception of the true 

 meaning of the latter. It may be granted willingly that the two 

 terms are very similar, and that it was quite wicked of anybody ever 

 to give such similar names to two distinct elements of the crinoid 

 skeleton ; but this does not alter the fact that the elements are 

 distinct. The confusion occurs in the following sentences. On 

 p. 904 it is written, " In the apical system [of the Echinoderma] the 

 following plates occur: (1) a central plate at the apical pole; (2) a 

 circlet of five radially-placed plates, the Infrabasals," etc. From 

 p. 888 we learn, quite correctly, that this central apical plate is called 

 the " Dorsocentral." On p. 915 follows this remarkable statement: 

 " A central plate has been observed in the larva of Antedon. Later 

 on it fuses with the uppermost ossicle of the larval stem, which 

 supports the calyx, and with the infrabasals, to form the Centro- 

 dorsal." That is to say, according to Dr. Lang, the dorsocentral 

 forms part of the centrodorsal. The facts are simple ; they are not 

 as above stated, but they are as follow. The plate in the Antedon 

 larva, to which the name " dorsocentral " has been applied, lies at 

 the distal or aboral end of the stem, and differs from the other 

 calcareous elements of the stem in being larger and flatter. It is by 

 this plate that the so-called " pentacrinoid " larva becomes attached. 

 When the larva begins to assume the characters of the adult Antedon, 

 the proximal ossicle of the stem, nearest the calyx, increases in thick- 

 ness, and develops first one, then two and three, whorls of cirri ; it 

 also becomes fused, as is now known, with the remains of the infra- 

 basals. The Antedon eventually breaks away from its stem, carrying 

 with it only this proximal, cirrus-bearing ossicle ; and it is to this 

 ossicle that the name " centrodorsal " is properly restricted. It is 

 obvious, then, that the dorsocentral has been left behind in the place 

 to which it became attached, and that it has no part or lot in the 

 centrodorsal. 



The centrodorsal is considered to represent more than one stem- 

 ossicle, as has been most clearly put by Ludwig (Zur Anatomie des 

 Rhizocrinus Lofotensis ; Zeitschr. f. wiss. Zool. xxix., p. 73, 1877) : 

 " The centrodorsal is a compressed upper portion of the stem, in 

 which the calcified tissue has undergone no separation into ossicles 

 lying one under the other." The fact that the infrabasals become 

 fused to this structure is interesting, but cannot be considered to con- 

 flict with the statement of Ludwig — that is to say, this morphological 

 element would remain a centrodorsal even if the infrabasals did not 

 become so fused ; indeed, it is possible that they were not so fused in 

 many of the earlier Antedonidae. The fact that the centrodorsal 

 bears whorls of cirri is, however, very characteristic, for it is chiefly 

 this fact that shows it to be a compound structure and not one simple 

 stem-ossicle. The early stages in the evolution of a centrodorsal may 

 be seen in Millericnnus Pratti, as described by P. H. Carpenter 

 {Quart. Jour. Geol. Soc, vol. xxxviii., p. 21), and in Thiolliericrinus (De 

 Loriol : " Faune jurassique du Portugal "), and this palaeontological 

 evidence confirms the view that the centrodorsal is a compound 

 structure. Essentially it is a part of the stem, and though it may, in 

 certain cases, become fused with some elements of the calyx, it 



