lowish-red colour. Notwithstanding repeated endeavours, I cannot succeed at this season of the year when the hens are moulting, in subjecting eggs to incubation for so long a period, I can, therefore, only give a representation of these vessels from a recollection of what I observed in the early part of this year. (See pl. IV, fig. 12.) In some situations the capillaries are perfect, and connected with the larger vessels; at others they have the appearance represented in the figure, and illustrated previously by observations on the tail of the tadpole. In addition to these capillaries, which form a network of canals of irregular caliber and give off blind branches, some separate irregular corpuscles are seen, such as h and i, which do not appear to be connected with the vascular network. These bodies send off blind processes of various forms in different directions, and have the appearance, therefore, of stellate cells. They have a yellowish-red colour, like that of the bone-capillaries, which circumstance is alone sufficient to suggest the supposition that they are cells of capillary vessels in progress of development. This becomes much more probable, when we observe some of these corpuscles, such as k, already connected with the true capillaries. We may, therefore, with a high degree of probability at least, regard them as the primary cells of capillary vessels; and in that case the description of the formation of these vessels, previously given, would be the correct one. The following would, therefore, be the mode in which the formation of the capillaries and of the blood takes place in the germinal membrane: among the cells which compose the germinal membrane, some which are deposited at certain distances from one another, are developed into the primary cells of capillary vessels by becoming elongated on different sides so as to form stellate cells. The processes of the different cells come into contact and coalesce, the septa are absorbed, and in this manner a network of canals of very irregular caliber is produced, the prolongations of the primary cells being much thinner than the bodies of the cells. These processes of the cells or passages of communication undergo expansion until they and the bodies of the cells all attain one equal width, until, in fact, a network of canals of uniform caliber is formed. The fluid portion of the
