328 PKOCEEDINGS OF THE GEOLOGICAL SOCIETY. 



and, in consequence, the ciphers of the antepenultimate and penul- 

 timate, being the two posterior intermediate molars, have heretofore 

 been but vaguely ascertained, Cuvier had not advanced sufficiently 

 far in the investigation of the subject to attempt to ascertain them. 

 De Blainville attributes from fifteen to sixteen ridges to the antepe- 

 nultimate, and eighteen or nineteen to the penultimate, in the upper 

 jaw. Owen considers the antepenultimate (fourth in succession) to 

 have been subject to considerable variation in the number and pro- 

 portion of the ridges, which he estimates as ranging from twelve to 

 sixteen, the greater number being usually in the lower molar. Of 

 the penultimate he describes the ridges as ranging even from sixteen 

 to twenty-four. Upon the examination and comparison of a very 

 large number of specimens I have been led to the conviction that, 

 ordinarily, the antepenultimate upper true molar repeats the duo- 

 denary cipher of the last milk-molar, and that the penultimate, as 

 in the Indian Elephant, advances by an increment of four ridges. 



First, in regard of the antepenultimate upper, or fourth in the 

 order of horizontal succession. A very fine illustration of this tooth 

 in situ is presented by a specimen in the Museum of the College of 

 Surgeons (no. 620, Cat. Poss, Mam. p. 153), comprising the palate 

 with a molar on either side, and in front of it the empty fang-pits 

 of the last milk-molar, whcih had been shed. The crown of the 

 antepenultimate is worn to the last ridge, but quite perfect, and 

 presents the disks of twelve principal ridges, with talons ; it is very 

 broad in relation to the length, and when compared with the corre- 

 sponding tooth of the existing Indian Elephant it looks short and 

 squat ; the outline is nearly a parallelogram, of which the length 

 is less than twice the width ; the disks of wear are closely approxi- 

 mated, forming narrow transverse bands ; the enamel plates are 

 very thin, with a slight tendency to minute irregular undulation, 

 nowhere amounting to crimping. 



The dimensions are : — inches. 



Length of crown 5-1 



Width of crown in front 2*4 



„ „ behind 2*4 



Greatest vddth of crown 2-75 



This specimen is of North American origin. 



De Blainville remarks that the penultimate upper (or fifth in the 

 order of succession) in the Mammoth is rare in the French collec- 

 tions. He was unable to include a figure of it in the rcih series of 

 representations contained in the ' Osteographie.' In the descrip- 

 tive details of the dentition (p. 189) he cites, as a fine illustration of 

 it, a specimen from Warsaw on the Yistula, having a crown still 

 composed of eighteen or nineteen ridges, although the most advanced 

 of these are worn out ; and he states that the tooth was remarkable 

 for its large size. These circumstances throw great doubt upon the 

 numerical rank assigned to it, which is strengthened by the fact that, 

 in the references to the plates (p. 357), De Blainville mentions that 

 he had no illustration of the penultimate except a bad cast, and 

 that it was therefore omitted. The Warsaw specimen is probably a 



