Development, &e., of Blood-vessels. By G. and F. E. Roggan. 575 



other, as stated by Golubew, whether the junction is effected at 

 right angles, as in Figs. 1 and 2, or in direct Hnear continuation, as 

 in Figs. 6 and 7. 



Hitherto we have been principally engaged with cells forming 

 attachments at or from the perij)hery to blood-vessels. Let us 

 now study those cases where the cells appear to be attached to or 

 at the blood-vessels, and stretch out from them to form a junction 

 with cells lying unattached. Cell a, Fig. 9, seems to offer a 

 suitable example where the cell seems to make the capillary its 

 base of operation, from which it stretches to form an angular 

 junction with cell d, the peripheral point of another capillary. 

 But on the other hand, cell a, Fig. 9, may be held to form the 

 next stage to cell a, Fig. 1, supposed to be going in an opposite 

 direction, that is to say from the periphery to the blood-vessel 

 centrally. In other words, it is difficult or impossible to determine 

 whether cell a, Fig. 1, should follow cell a, Fig. 9, as a type of 

 cells acting or passing from the capillary, or cell a, Fig. 9, ought 

 to follow cell a, Fig. 1, as a type of cells passing to the capillary. 

 We may leave the question undecided, for it does not really much 

 matter, and it is only useful in serving to reconcile the opposing 

 views of Kolliker and Golubew as to whether it was by a cell 

 passing to or passing from the capillary that new vascular exten- 

 sions were formed. 



If, however, we still follow the process in the sense that cell «, 

 Fig. 9, is a type of cell acting at or from the blood-vessels, we may 

 find an undoubted example of the same principle in cell e, Fig. 8, 

 which, while lying upon the capillary, has begun to stretch out a 

 short fine process of its protoplasm towards a wandering cell, /, of a 

 circular form, which has not yet begun to elongate its protoplasm 

 into the invariable bipolar shape which characterizes the wandering 

 cell when it has undertaken the duty of a vessel-forming cell. 

 Cells m and n, Fig. 12, arc very good examples of this direction of 

 development, and cell h is even more typical, because it has not yet 

 begun to send out any process peripherally. It must, therefore, 

 appear abundantly evident that the process of prolongation of 

 newly forming blood-vessels by cells may bo either towards or 

 from the vessels. We have shown that, although cell a in Figs. 1, 

 2, and 9, and i, Fig. 12, may be on debatable ground, yet such 

 examples as cells a and e, Fig. 8, are undoubtedly extreme examples 

 respectively of direction of growth from and towards the blood- 

 vessel. 



{Sometimes we have cells connecting the blood-vessels before the 

 process of vacuolation has begun in them, as in Fig. 10, and again 

 we may lind earlier stages than that seen in Fig. 1, as for example 

 in cell/, l''ig. 0, which is evidently only approaching the capillary 

 in course of formation, and lying at right angles to the joint 



