204 Egbert Thompson Young, 



cell, f. The incompletely developed capillary, c' is very evidently 

 being foi'med in the parenchyma strand which is a process of the 

 cells cc' and f. ^) Fig-. 40 shows the same mode of development 

 Here one finds a group of four developing flame cells, mf with their 

 funnels in connection with the excretory duct, le thru the medium 

 of the protoplasmic strand, c which, with its cell, cc forms the 

 rudiment of the future capillary. The development of the funnel 

 from the flame cell is evidenced by the origin of the ciliate process 

 within the cell (Figs. 21, 28 etc.). That part of the cell surrounding 

 the ciliate process is unquestionably destined to form the funnel. 



The development of the flame cell precedes that of the capil- 

 laries. The latter may develop in connection with a capillary cell 

 (the so-called "fourth cell" of Bugge), as in the above-mentioned 

 figures, or it may develop from the processes of the fiâmes directly, 

 without any connection with a capillary cell (see Fig. 42). While 

 the capillary is fundamentally nothing more than an ordinary paren- 

 chyma strand in which a lumen is formed, and its cell a modified 

 parenchyma cell, still it is evident from the great ability of the 

 capillaries to bud and anastomose and thereby assume such an ir- 

 regularly twisted and devious course as they do thru the parenchyma 

 (Figs. 38 and 42), that the capillaries with their cells are specialized 

 structures and not merely chance passages in the parenchyma. This 

 statement, while apparently contradictory to the one above, in which 

 I claim this simple origin for the capillaries, is not actually so. At 

 first the capillaries are simple intra-cellular spaces in the paren- 

 chyma, but later they become specially modified structures. 



I have never yet discovered more than one cell to a single 

 capillary, but any one who has attempted to trace the course of one 

 of these structures in all its devious windings will realize the 

 practical impossibility of saying a final word on this subject. It is 

 evident, however, that a part of the capillary belongs to the flame 

 cell or cells and therefore it may be said to have a multicellular 

 origin. 



Occasionally capillaries are developed without any connection 

 with a flame cell (Fig. 36), but it is probable that such a connection 

 will later be established. In Fig. 38, the connection of the 

 capillary, c', with the cytoplasm about the nucleus of the flame cell, /", 



1) It is interesting to note in this instance the independence of 

 capillary and funnel. 



