404 
ME. S. J. A. SALTEE ON THE STEUCTTJEE AND 
It is sometimes very difficult to discover why the detached particles thus maintain 
their mutual relationship, while at others one is able distinctly to see a film of extreme 
tenuity holding them together. It is almost impossible to see this film as long as these 
particles are adherent to and resting upon the plates ; but when detached, it may gene- 
rally be made out, and it is especially intelligible when it projects beyond the margin 
of one of the particles (Plate VIII. figs. 4 & 5, d). 
I had long much difficulty in determining the precise character of this film, but I am 
now entirely satisfied that it is of the same chemical nature as the particles themselves, 
and that it is double, passing from each face of each particle to each face of the con- 
tiguous ones ; thus converting the intervals between them into canals, and constituting, 
when thus viewed in face, an upper and a lower wall to those canals. Now, though 
this film is of the same nature as the plates themselves, and ultimately becomes indis- 
solubly united with them, still, as long as the particles can be detached en masse and in 
unbroken relationship, -with the film entire, it cannot be said that the intervals between 
them are mere interstitial vacuities destitute of walls, as Professor Williamson asserts; 
the truth being that the tubular system is, by means of the soldering particles and their 
connecting films, introduced among — interjyolated hetiveen — the other previously-existing 
elements of the tooth, and that it has an existence, as a, tubular system, before the indis- 
soluble adhesion of the elements has occurred. 
How early the soldering particles are held together by the calcareous film that ulti- 
mately connects them, I am unable to say ; but as far as observation goes the particles 
appear quite separate at first. Neither can I speak with certainty of the universality of 
this connecting medium, though from its usual appearance, when properly sought for, I 
should infer that it is always, or at least generally, formed in the advanced state of the 
tooth’s development. 
The soldering particles themselves vary in diameter indefinitely from the minutest 
microscopic point to a disc of the x-^th of an inch. Their average may perhaps he 
stated at the iMToth or xwo^h of an inch, and their thickness from one-fourth to one- 
sixth of their diameter. 
Their form is usually circular or oval at first, but as they increase in size and approach 
each other they become polyhedral and angular. Each soldering particle has a m.ore 
or less conspicuous nucleus, around which are seen series of concentric rings — incre- 
mental lines of progressive enlargement. Contrary to Professor Williamson’s state- 
ment, the neighbouring particles frequently fuse together, and their original distinct 
centres are easily recognized (see Plate VI. fig. 5, d, c) ; indeed it is by this complete 
and general fusion in the centre of the tooth that the compact hard axis is produced in 
which no interspaces remain. 
Thus is built up this curiously complex and, at the same time, definitely planned 
fabric. It has many points in common with all the skeleton structures of the Echino- 
dermata — the repetition of similar primary elements, and the sprouting from their sur- 
face of secondary elementary parts, and the further progi’essive enlargement of all by 
