50 Sir Philip Grey-Egertdn — On some New Pycnodonts. 
row characteristic of the larger teeth, but it occurs in tlie centre of 
the denticle, and is deeper in proportion to the size of the tooth, so 
that it is hardly altogether effaced by use. The four posterior teeth 
of the third row (PL III. Fig. 1 c) are smaller than those of the second, 
and the outline is less elliptic and more irregulär. These also have 
the central furrow described before, but modified in harmony with 
the form of the teeth. The teeth in front of these (PI. III. Fig. 1 d) 
are in a double row, and seem to have arisen from divided germs, 
each pair making up the form and size of one of the posterior teeth, 
suggesting the idea that the single teeth of the hinder portion of the 
rank are replaced by twin teeth in the front of the mandible. This 
peculiarity will be alluded to more particularly hereafter. The 
result is a great irregularity in this part of the dentition. Some of 
these small teeth are elliptical, some more or less rounded, some 
trigonal ; some have the larger axis to the front, others coincide 
with the two inner ranks in having the larger axis transversal. 
They all, however, have a central pit varying in form with the out- 
line of the tooth. The mandibular bone (PI. III. Fig. 1 m) extends 
an inch or more backwards beyond the dentigerous area ; it is close- 
grained and solid, and measures an inch in thickness in the middle, 
and thins off a little towards the symphysis, and more so to form the 
outer edge of the mouth. 
Before attempting to determine to which genus of the Pycnodont 
family this specimen is referable, it will be necessary to notice 
some of the discrepancies which occur in the writings of those authors 
who have treated of this subject with regard to the number of the 
ranks of mandibular teeth. Agassiz 1 says, “ The lower jaw is car- 
petted with large teeth arranged in three orfive ranks on each side.” 
Pictet 2 follows this dictum. Wagner and Thiolliere, 3 on the contrary, 
maintain that the correct number is four ; but in the figure given by 
the latter of Pycnodus Bernardi (pl. 5, fig. 2), five rows are distinctly 
shown. Sauvage 4 divides the genus into those with four rows and 
those with more. The great opportunities afforded me of examining 
these interesting remains of a family entirely extinct since the 
Miocene age have led me to think that much irregularity obtains in 
the development of the marginal rows of tritoral teeth, and that 
these variations are not of specific value except in cases where they 
are constant. In Pycnodus Bucklandi, JTugii and parvus there are 
always five and sometimes six- rows. The inner row when present 
is invariably composed of small teeth, which are sometimes irregulär, 
sometimes have fallen away through use or age, and in one specimen 
in my collection are duplicated. The second row is always com- 
posed of the largest teeth and is generally constant in character. 
The third row is also for the most part regulär, but composed of 
much smaller teeth than the second. The outer rows are subject to 
great irregularities in number, form, and position. In some cases 
1 Poissons Fossiles, vol. 2, pt. 2, page 183. 
2 Traite de Paleontologie, vol. 2, p. 197. 
3 Poissons Fossiles du Bugey, p. 1 1 . 
1 Poissons des Formations Secondaires du Boulonnais. 
