August 22, 1884.] 



SCIENCE. 



143 



such matters, or the exhibiting departments of 

 the government will be forced to the necessity 

 (to do proper credit to themselves) of main- 

 taining exhibition series, which, with slight 

 modifications for special occasions, may be 

 kept at hand, to send wherever and whenever 

 required. If we are rightly informed, the na- 

 tional museum has already decided on some 

 such step ; and, if international exhibitions are 

 to be a yearly occurrence, the museum should 

 add to its staff a special exhibitionary force, 

 and not weaken its efficiency for its proper 

 work by these constant extra draughts upon 

 its energy. 



LETTERS TO THE EDITOR. 

 Classification of the Mollusca. 



In Mr. Dall's kindly notice of the article 'Mol- 

 lusca' in the 'Encyclopaedia Britannica,' published 

 in your journal of June 13, he attributes to me " the 

 erroneous statement that the radula of Glossophora 

 is horny," and adds that 'it is really chitinous.' In 

 the ordinary sense of the word 'horny,' chitin is (I 

 venture to think) correctly described as horny. That 

 the radula is generally considered to consist of the 

 chemical body known as chitin is distinctly stated in 

 the article criticised by Mr. Dall. At the bottom of 

 p. 460 occur the words, 'a chitinous band (the rad- 

 ula).' I should be glad to know if Mr. Dall has 

 undertaken any special chemical analysis of the sub- 

 stance of the radula (1). 



With regard to the very general presence of jaws 

 in glossophorous Mollusca, I must maintain my state- 

 ment. The presence of a calcareous impregnation is, 

 it is true, not usual, but exceptional (2). 



Mr. Dall is mistaken in supposing that I have fol- 

 lowed Macdonald in regard to formulae for the teeth 

 of the radula. The other writers whom he cites as 

 not followed are precisely those from whom my state- 

 ments on the details of this subject were drawn (3). 

 I have no fault to find with Mr. Dall for differing 

 from me as to certain points of classification, but I 

 should be glad to know his grounds for regarding the 

 Zygobranchia as an artificial group. He merely re- 

 asserts the old view, which I think I have sufficiently 

 shown to be untenable (4). Mr. Dall also asserts 

 that the orders of Lipocephala, based on the charac- 

 ters of the adductor muscles, are defunct. In spite of 

 this opinion, the muscles themselves still exist, and, 

 in my opinion, furnish indications of natural and 

 important divergent groups among the bivalves (5). 



I should be glad to know on what grounds Mr. Dall 

 considers the three divisions of Lipocephala adopted 

 by me to be unnatural. 



Lastly, let me say that I do not know on what 

 authority Mr. Dall asserts that the calcareous devel- 

 opments of the integument in Chaetoderma and Neo- 

 menia have no relation to the shells of Chiton. That 

 they also represent or replace the spines of Chitons 

 is sufficiently obvious. But what is to prevent our 

 conceiving of the epidermic shelly plate of a Chiton 

 as originally developed by the gradual coalescence of 

 a number of small calcareous denticles, in the same 



way as the mesodermic dermal bones of bony fishes 

 have developed from the shagreen denticles of the 

 sharks (6) ? E. Ray Lankestek. 



University college, London, 

 July 23. 



(1) Not being an organic chemist, I have not 

 attempted analyses, but have tested many radulae 

 with one result, — the cutting points of the teeth 

 are always, and the whole radula generally, of a sub- 

 stance allied to chitin. The very generally erroneous 

 statements in the text-books led to the criticism of the 

 language of Professor Lankester as tending to con- 

 tinue the confusion. Chitin is surely as different 

 chemically from horn as bone is, and it cannot be 

 desirable to continue to treat the two substances in a 

 way to perpetuate an error. Further data on this 

 topic may be found in the August JVaturalist, pp. 

 776-778. 



(2) I should be grateful to Professor Lankester 

 for the name of any recent mollusk having a ' shelly ' 

 or even a partially ' calcified ' jaw. 



(3) The formulae given for the teeth, and the 

 method used in making a formula, as inferred from 

 the text, which were the particular details criticised, 

 are partly incorrect. I was wrong, however, in 

 assigning a source to them. One (for instance, 

 Patella vulgata) has the formula 3+3+1+3+3, in- 

 stead of 3.1.4.1.3. No mollusk has more than one 

 median tooth; and the central figure of the formula 

 must in all cases be 1 or 0. I find the erroneous 

 formula in Sars's text, though he figures the teeth 

 correctly. Again : Chiton stelleri has, like all Chitons 

 hitherto examined, the formula 6+2+1+2+6, instead 

 of 0000.1.1.1.0000, which is given; but this is doubt- 

 less copied from some other authority. However, 

 accurate formulae for the Chitons and Limpets have 

 been accessible for some years. Again: the teeth of 

 the radula are divided by nearly all modern students 

 of that organ into rhachidian or median, lateral, and 

 uncinal teeth, — three series which have anatomical 

 relations to the radula, which are usually pretty clear. 

 For ' lateral ' Professor Lankester substitutes the 

 term ' admedian,' which is not, as far as I know, 

 in use ; and for the ' uncini ' he adopts the term 

 ' laterals,' which I venture to think is undesirable 

 as leading to confusion, and not in accord with 

 general usage. 



(4) The grounds on which I sustain the generally 

 accepted views of malacologists, as to the relations 

 of the groups Professor Lankester has compounded 

 into the order Zygobranchia, are, that the mere abor- 

 tion of one of a pair of organs is not a character of 

 ordinal value; nor are the characters assigned to 

 Zygobranchia applicable to all its members. More- 

 over, I am of the opinion that the characters which 

 unite the Rhipidoglossa among themselves and the 

 Docoglossa among themselves are of higher systematic 

 value than the characters here relied upon for dis- 

 membering them. I believe, that, had the learned pro- 

 fessor made researches among a large number of these 

 forms, he would probably be of this opinion also. 



(5) The characters of the adductor muscles, as long 

 as we were ignorant of intermediate forms, seemed 

 to afford a good basis for orders in the Lipocephala. 

 Now that we know of forms which are more or less 

 intermediate, in the Pectinidae, Ostraeidae, Mytilidae, 

 and other families, and that in the young (not embry- 

 onic) there are frequently two adductors discernible 

 in supposed monomyarians, with such forms turn- 

 ing up as Dimya, and, more recently, Chlamydo- 

 concha, all tending to efface the supposed definite 

 limits between the alleged orders, it seems impos- 



