574 Transactions of the Society. 



of vacuolation become larger than in i, Fig. 12, and its junction 

 with the capillary strengthened by the addition of two cells h 

 and c, but already at its peripheral pole it has formed a connec- 

 tion in direct linear series with another bipolar cell, d, which has 

 arrived at the same degree of development as a in Fig, 1. We 

 shall hereafter refer to the bodies seen within the vacuole ; in the 

 meantime we have to notice that the vacuole has not become con- 

 tinuous with the cavity of the capillary, but this last phase is seen 

 between a and 6, Fig. 11, where the cavity is continuous with the 

 lumina of the vessels to which they are attached. 



So far we have traced the history of the development of a 

 wandering cell into the first link of a developing blood-vessel. The 

 formation of the second link includes the process of prolongation of 

 a blood-vessel by a cell in direct linear continuation. Cell d in 

 Figs, 3 and 4, represents the first stage, in which nothing particular 

 is to be remarked beyond what we have already described. But in 

 d, Fig. 5, we have a stage further advanced ; a cavity has already 

 vacuolated in the direction of a, the cell also vacuolating to which 

 it is attached on the side next to the blood-vessel. 



We may here call attention to an interesting peculiarity gene- 

 rally observed in cells vacuolating to form blood-vessels. Cell d, 

 having as yet no cell on its peripheral end or pole, has thrown 

 the whole of its protoplasma into the duty of forming an attach- 

 ment with cell a on its central aspect; and for the same reason 

 the cavity of vacuolation is formed on the central side of the 

 cell nucleus, while cell a, which wishes to form a connection cen- 

 trally with h and peripherally with d, has vacuolated at both sides 

 or ends of its nucleus. Cell i, Fig. 12, showed the earliest stage 

 in the carrying out of this principle, and we may see the third 

 stage in cell d, Fig. 6, whose vacuolated cavity at the attached 

 side of its nucleus has formed a junction with the cavity of the cell 

 a, to which it is attached centrally, and which, having vacuolated 

 before forming any attachment to cell h, shows the vacuolation 

 only on that side of the nucleus nearest to its peripherally attached 

 neighbour cell d. 



In Fig. 7 we have the same process another stage further on, 

 where not only have the vacuolated cavities in a and d become 

 connected, but they have gone on extending themselves beyond 

 their respective nuclei both centrally and peripherally ; and further, 

 the oblique splice or union between a and d has been strengthened 

 at its weakest point by the addition of cell c placed according to the 

 usual rule. In this figure the dotted line represents the continua- 

 tion of the black silvered line that marks the union of the two cells 

 a and d on the opposite surface of the tube. In all these examples 

 it will be observed that we never have cells joining on the end-to- 

 end principle, as stated by Arnold, but, as a rule, they overlap each 



