DEVELOPMENT OF VASCULAR DENTINE. 
33 
tubes become fewer and fewer, irregular in course, and finally one third of the way 
down altogether cease (see fig. 10*). But the lower portion of the dentine which is 
devoid of dentinal tubes presents the longitudinal striation, and is, in fact, precisely 
like the dentine of the Gadidse. And at the point where the dentinal tubes are few 
in number (t in fig. 10*), the longitudinal striation is traversed by transverse dentinal 
tubes, thus proving by the co-existence of the two that they are due to different and 
independent causes. 
I cannot therefore confirm the opinion of Retzius and Professor Owen, that the 
whole tooth substance of the Gadidse is permeated by a minute tube system : on the 
contrary, I believe it to be quite solid. 
But although it is solid the matrix of the dentine is not quite homogeneous ; in 
transverse section very faint striae radiate outwards from the pulp cavity, and the 
spaces between the striae (the distinctness of which has been intentionally slightly 
exaggerated in fig. 7) are mapped into a finely reticulate pattern. 
Under a higher magnifying power this pattern is found to present the appearance 
shown in fig. 3, and as this appearance is found alike in transverse, longitudinal, and 
oblique sections, it is probable that it is due to calcification first taking place in such 
manner as to form isolated globules (as indeed happens to a greater or less extent in all 
varieties of dentine), and to these globules failing to completely coalesce but becoming 
modified in form by mutual apposition. 
That that is the true interpretation of the nature of the pattern is indicated by the 
occasional occurrence of globules which have not been thus distorted, but retain their 
spherical form, and also by the appearance of fine reticulation in the dentine of the 
teeth of allied fish, such as the Pleuronectidse. 
The pulp is less rich in blood-vessels and far more rich in connective tissue than 
that of the Hake ; otherwise, it does not call for any special description. On its 
surface the odontoblast cells form a very distinct layer about - 5 -g-g- of an inch in thickness, 
and capillary vessels may be seen running out from the pulp through the odontoblast 
layer into the dentine (see fig. 9). But I have never seen capillaries clothed with 
odontoblasts like that of the Hake represented in fig. 7, though this may be due to my 
never having examined a Cod-fish so freshly caught as was the Hake from which fig. 7 
and fig. 2 were drawn. 
The dentine in the Haddock, Whiting, Coalfish ( Gadus carbonarius), Pollack, and 
Ling, does not differ in essential particulars from that of the Cod. The frequency, size, 
and the form of the capillary loops vary, so that anyone well familiarised with their 
respective appearances might probably succeed in identifying them, but the differences 
are in minor points, and the structure of the matrix is similar in all. 
In all of them we have the same essential features, the penetration of actual hard 
dentine by capillary vessels, and the absence of true dentinal tubes. 
But although I have not found the dentinal tubes in any of the Gadidse which I have 
examined, and although their absence is the rule in this modification of dentine, no 
MDCCCLXXVIII, F 
