The development of the skeleton in the Genus Waldheimia. 381 
ed to the hingeplate by the erura, secondly, the lamellæ are 
connected with a septum, and finally the reflected part of the 
loop is connected with the lamellæ and the septum by two 
vertical walls placed close together; as the shell is enlarged 
the loop expands in breadth; the united lameelle split from below 
upward, dissolve connection with the septum and the lateral 
walls vanish. In this representation no correction is to be made, 
- but beside being now enabled to proceed one step further into the 
development, I can also replace my earlier simple drawings by 
better and more complete illustrations. These are all drawn by 
camera lucida. 
The earliest stage at which a coherent apophysary system 
of the Waldheimia cranium is observed, has a size a little 
less than 3 m. m., the hoemale valve being 2 m. m. Fig. 1: 
Two long thin lamellæ project from the erura, connect with a 
filiform septum and run together in an acute angle in front, 
where they unite. By the connection of the lamelle two close- 
set walls are given off, which by a reflection form a tube, the 
posterior end of which is closed. In Fig. 2 a. b. the size of the 
hoemale valve has reached, 2,5 m. m., but the state of things 
remains the same, save that both the loop (or the tube) and 
the lamellae have expanded. 
Fig. 2° a the lamellæ, 6 the two from the lamelle issuing 
vertical walls, or the lateral walls of the tube, ce the closed tube 
(or the loop). 
The first visible change occurs by an opening in the closed 
end of the tube (Fig. 3 a, b). The united lamellæ then begin 
to split apart at the anterior end (Fig. 4). The hoemale 
valve has now attained the size of 4,5 m. m. A continous ex- 
pansion of the loop-complex causes a perforation in the lateral 
walls (Fig. 5), and the septum becomes thinner and thinner. 
The next sizes observed were 5,2 m. m., and 5,6 m. m. (hoe- 
male valve). Figs. 6 and 7: The same characteristics as those 
ense in Fig. 5 are still in the main prevailing, but the con- 
Archiv for Mathematik og Naturvidenskab. 26 
