135 



and many other shells. They are in the greatest profusion at the ex- 

 panded end of the fan-shaped plate, and, towards the extreme edge, 

 are so thickly disposed, and so tortuous in their course, as to render 

 it exceedingly difficult to follow their lines of direction with the eye. 

 At a short distance from the expanded edge of the plate, towards the 

 pointed base, they are less tortuous in their course, are loosely 

 grouped in bundles of three or four together, and approximate to- 

 wards straight or slightly curved lines, as represented at PI. xvi. 

 fig. 2, a a. They are not continued towards the pointed base more 

 than a fifth or a fourth of the length of the plate, but usually termi- 

 nate abruptly in a ccecoid form, without the slightest previous dimi- 

 nution of their diameters. They do not often divide or anastomose, 

 excepting near the expanded end of the fan, but at that part of the 

 plate these occurrences are frequent. In the spaces between the 

 bundles of canals, in some of the plates, near the broad end, there 

 are numerous detached groups of convoluted canals, apparently of 

 the same description of organ as the longitudinal bundles, but not 

 emanating from or in any manner connected with them; and in other 

 plates I found another modification of these curious organs, in the 

 form of numerous, short, straight canals, which are disposed at right 

 angles to the longitudinal ones, and, like the convoluted variety, have 

 no connexion with them, as represented at PI. xvi. fig. 2, b b. Up- 

 on carefully measuring the canals, I found the longitudinal ones 

 ■ r4 | B a of an inch in diameter, and the short transverse ones ar ^ oa . 

 In Unto, when fractured at right angles to the natural planes of the 

 shell, we find immediately beneath the periostracum a thin stratum 

 of large cells, having their axes at right angles to the natural surfaces. 

 Immediately beneath these is a series of thin plates, which pass back- 

 wards in a diagonal direction towards the inner surface of the shell, 

 where they terminate. Each of these plates is composed of an infi- 

 nite number of polygonal tessellated cells, filled with carbonate of 

 lime. The same tessellated cellular structure prevails in Mytilus. 

 The cytoblasts may be occasionally observed in the detached cells of 

 both shells, and especially in the latter, when viewed by transmitted 

 light, with a power of 1000 linear. The shells from which I obtained 

 the tessellated cells in the greatest perfection, were the fossil Unios 

 from the fresh-water formation of Grays, in Essex, and from speci- 

 mens of Mytilus found in a very recent diluvial deposit at Brackle- 

 sham, Sussex. In these the animal matter has been decomposed to 

 such an extent as to leave them in an exceedingly fragile state, and 

 it is from specimens in such a condition that the best idea of the 



L 



