352 University of California Publications in Zoology [Vol. 19 



body one would certainly say that each joined the collecting tubule or 

 its accessory singly instead of in threes. However, knowing of this 

 grouping into threes at the posterior end, it is not difficult to make 

 the proper association here. This widening of the distance between 

 the attachment of the capillaries in a group is accounted for by the 

 gradual loss of the accessory collecting tubules or by the merging into 

 one of this and the main collecting tubule. 



These groups of three flame cells are arranged dorsally and ven- 

 trally in a definite way. The most anterior group is essentially a 

 dorsal group appearing quite close to the upper surface. The next 

 group of threes is a ventral one, the next one is dorsal, the next ventral, 

 and so on to the posterior end. I cannot be absolutely sure of the 

 dorsal-ventral arrangement of the last two groups, however, because 

 the groups are much coiled, yet I am reasonably certain that they are 

 so arranged. This arrangement is probable if the entire body is to be 

 drained equally. Again within the group of threes there is a definite 

 arrangement. Starting at the anterior end the first flame cell is dorsal, 

 the next ventral, the next dorsal ; then in the next group of threes the 

 first is ventral, the next dorsal, the next ventral, and so on. This again 

 is a proper arrangement, if dorsally and ventrally the body is to be 

 equally well drained of wastes. These two plans, or some modification 

 of them. I should expect to find in all echinostome cercariae and adults. 



Prom a study of the literature and the drawings of the excretory 

 systems of cercariae that have been worked out, and from my own 

 work, it seems probable that the flame cells and their capillaries are 

 arranged in definite groups connecting with the accessory collecting 

 tubules, and that each group though modified through increase, reduc- 

 tion, or loss of parts has come from the mitotic division of a single 

 flame cell. In some cercariae, such as those of Schistosoma japonicum 

 or Cercaria douthdtti, the capillary groups have not yet arisen, since 

 the capillaries are arranged singly, but I should expect to find in the 

 adult worm each flame cell in the cercaria represented by a group of 

 flame cells. In Echinostoma revolutum the groups are already present, 

 hence in the adult I should expect to find the same number of flame 

 cells. The excretory system of the cercaria of Echinostom.a revolutum 

 has been modified somewhat and is hard to interpret as a definite 

 unified arrangement throughout. This theory of the formation of 

 capillarly groups was first given by Looss (1894) and later accepted 

 by Cort (1918a). Even in 1881 (pi. 1. fig. 4) Fraipont shows the 

 formation of two flame cells by mitotic division. 



