510 EMBRYOLOGY OF THE TURTLE. Part III. 
sharp, (Pl. 9a, fig. 48a,) when seen from the same point of view; and in this 
case the outline of each nodule is serrated irregularly by the projecting ends of 
the crystals. 
When a section is made through the thickness of a nodule, its centre, at the 
base, (PI. 9a, fig: 
2. 43, 6,) is seen to be the point from which all the radiating lines 
proceed to the surface, and the concentric lines, in the view from above, (fig. 43a,) 
both at this age and also in the young nodules, (Pl. 9a, fig. 44a,) appear arched, 
showing that the lime crystals are arranged as if around a sphere, the centre of 
which coincides with the starting point (Pl. 9a, fig. 43, b) of the radiating lines, 
and that these radiating lines (fig. 43, a, 6, 43b) are the long sides of the pris- 
matic calcareous crystals. 
It will readily be seen, here, how air and moisture may gain access to the 
interior of the egg, when we consider that the nodules (PI. 9a, fig. 42, 43, a, a’) 
are not soldered to each other laterally. In the case of hard, brittle, and smooth 
shells, however, such as those of Xerobates, Platypeltis, Cmosternum, and Ozo- 
theca, (in which the groups of crystals become interlocked with each other by the 
dovetailing of their ends at the basal and younger portions of the nodules, and 
the later developed and more exterior parts of these nodules so confuse the termi- 
nations of their adjoiming crystals as to disguise their line of junction, and thus form 
a continuous stratum from one of the nodules to the other, like a universal bridge 
over the whole egg,) open spaces must be left in the lime deposit, m order that 
this may obtain, and we actually find it to be the case. In some species these 
spaces are quite numerous, as in Ozotheca (Pl. 9a, fig. 46); but in others they 
are more rare, for instance, in Platypeltis and Xerobates. 
From the remarks annexed to the table which we have given, (p. 508,) it will 
be seen, that, within each family of Turtles, the peculiarities of the egg shell are the 
same throughout. Thus, in the Chelonioidz it is nodular, and each nodule widely 
separate from its neighbors, and very ragged and friable; in Trionychide it has a 
continuous smooth surface bounding a uniform, dense, brittle stratum, which equals 
about one third the thickness of the nodules beneath; in Chelydroide it is nod- 
ular, (PI. 9a, fig. 42, 45, 45a,) and each nodule is smooth, hard, and brittle, and 
separable from its neighbor, although at the time of their formation they may 
sometimes be in contact with each other; in Cinosternoidx its continuous surface 
is wavy or pitted, terminating a uniform stratum, just like that of Trionychide, 
excepting that here it is about twice as thick, and fully two thirds the depth of 
the nodular part beneath; in Emydoide it is nodular, the nodules being similar in 
structure to those of the Chelydroid, but more closely united to each other, espe- 
cially in Emys Meleagris, (which by the way belongs to a distinct sub-family of Emy- 
doide,) where they are very closely set together, so that the shell is quite smooth 
