much as to be scarcely thicker than a fibre of areolar tissue (as in c). Branches are also sometimes given off from these wider parts, which likewise diminish very rapidly to the same degree of minuteness, without reaching another dilated part (as at d e), and which are, therefore, blind ones. According to the above view of the development of the capillaries, these appearances may be explained in the following manner: the wider portions, a, b, &c., are the bodies of the primary cells. Hollow processes, as at d, are sent out from the bodies of the cells as the result of a more vigorous growth in different situations, precisely as is the case in all stellate cells. These prolongations meet with similar ones from other cells, and thus produce the form c. But being hollow, they are capable of expansion during their growth, and thus the canal c becomes converted into f, and at length into g, which is as wide as an ordinary capillary vessel. A more accurate analysis of the observations, however, is necessary to enable us to judge of the correctness of this explanation. It might be doubted, in the first place, whether these were really capillaries. The blood flows uninterruptedly through the ordinary capillaries, but there are no blood-corpuscles in these canals, at least in the more minute ones; they are, therefore, more difficult to discover, and readily give rise to a doubt whether they are canals. But their direct continuity with the ordinary capillaries may be clearly demonstrated, and blood-corpuscles actually enter the wider ones. If they be true capillary vessels, they may either be ordinary ones in a state of contraction, or they must represent a certain stage of their development. But if it be difficult to conceive that a capillary vessel can have the power to contract itself almost to the minuteness of a filament of areolar tissue, such an assumption cannot be supported at all in respect to the blind branches, which do not join any other vessel, as at d. This form might, indeed, be admitted to be a certain stage of development, although not of the kind de- scribed above; but branches might be sent off from the capillaries already existing, which again might give off others. The objection, that such an explanation does not account for the varying width of these capillaries, might be met by assuming that circumstance to depend upon the surrounding substance. It is, therefore, necessary to see the primary cells
