30, 



illary ,- reaping cf a system cculd be progressively modified 

 until the axact pattern if? somewhat obscured* 



fhe drawing (text fig. He. 1) show* particularly what I 

 thin)' has happened to the accessory collecting tubule (3e), 

 to which ore attached the capillaries draining flame cells 



/en, eight and nine. Compare with figure 47 as it really 

 appears. As shown graphically this accessory collecting 

 tubul* ie quite long whereas in the actual specimen I 

 thin*" it i* reduced to a mere bulb t to which the capillaries 

 join* In cif?e f the accessory collecting tubule pb) shown 



to which are attached capillaries 4x, tx t and 

 in 

 we find that/ the drawing thin accessory collecting tubule 



ha* entirely disappeared or perhaps better is stretched 

 cut between thes oapillariea. the same condition exists 

 in case of the accessory collecting tubule of cmpillaires 

 ix, 2x and 3x. Hear the posterior end cf the ascending 

 tubular division of the bladder (1), are joined by collecting 

 tubules (2b and 2o), one on the left and the other on the 

 right hand side of each division of the V- ladder. The on* on 

 the right side (&b) receives, as shown graphically, two 

 accessory oclleoting tubuloe 3d, 5c, which in figure 47 are 

 entirely absent or are stretched out between ICx, llx, 12x f 

 and 13x 14x, and 15x respectively. The collecting tubule 

 n the left (2o) merges directly into *cce*sory collecting 

 tubule ($f) ( r>a shown graphically and in the drawing has ceased 

 to exist unlo^^ it is to be found between the capillaries 

 (17x and 18x) * Th last aclleoting tubule, (d), in the 

 graph and in the living specimen is nhort, to which join 

 'ory collecting tubules (3g and 3h). TPius is seems 

 I have suggested as the primitive pattern has been cm- 



