312 0. C. Farrington — The Nephrostomes of Bana. 



openings so minute could produce currents or accumulations of par- 

 ticles large enough to be seen from the surface. Applying the 

 experiment of Nussbaum, however, at seven different times to both 

 male and female frogs of various sizes, the evidences of ciliary action 

 previously described were observed in every case. 



in. Our belief that the nephrostomes in the frog open into the 

 blood capillaries of the kidney, rests mainly on the fact that sections 

 prepared after Nussbaum's method show that the carmine is present 

 not only in the nephrostomes, but in the blood capillaries and large 

 veins which are near the ventral surface of the kidney. This is shown 

 in both figs. 2 and 4. Applying the method of Nussbaum to the newt, 

 we find the cai*mine invariably in the urinif erous tubules and not in any 

 other part of the kidney. In the frog, however, the carmine is not 

 found in the tubules at all but only in the blood vessels. This makes 

 the evidence almost conclusive that the nephrostomes open into the 

 tubules in the one case and into the blood vessels in the other. We 

 need, however, to trace the tube directly to its termination before we 

 can assert positively its connection, since it may be j^ossible that the 

 carmine could make its way into the circulation by some other course. 

 Nussbaum states that by sections cut in glycerine he Avas able to fol- 

 low directly the course of the carmine from the nephrostome into the 

 blood vessel. It is very doubtful, however, if the evidence from 

 such sections could be relied upon, since the writer found that unless 

 some fixative for the carmine was iised the particles were scattered 

 by the knife through every part of the kidney. To overcome this 

 difliculty use was made of a method for imbedding proposed by Prof. 

 S. I. Smith. The kidney was imbedded in celloidin, but instead of 

 trving to make sections of this block by the usual method of cutting 

 under alcohol, the block Avas reimbedded in jiaraftine. From this 

 block, then, sections could be cut as usual with parafiine, and the 

 celloidin held all particles so firmly in place that there could be no 

 doubt that the place of the carmine in any part was that it had natu- 

 rally reached. From fig. 4, which represents a section prepared 

 after this method, it will be seen that the carmine appears only in 

 those blood vessels near the ventral surface. This was invariably 

 true, and hence the conclusion that the nephrostomes open into the 

 capillaries fiowing into the vena cava. This fact, moreover, makes 

 it quite unlikely that the carmine could have entered the circulation 

 in any other way than through the nej^hrostomes, for it would then 

 be equally distributed through the blood vessels of the kidney. To 

 test this point further, the vena cava inferior of a live frog was tied 



