120 Sir A. E. Wright. On I ntertr action" between 



fluid, descending in definite streams to spread itself out upon the floor of the 

 cell. Exactly the same applies to the heavier salt solution. It is not only 

 caught up into the serum, hut continues to ascend there in definite streams. 



Fig. 1. Fig. 2. Fig. .3. 



Fig. 1. — Flat diffusion cell filled in with 6 per cent. NaCl solution below. Upon this 

 is superposed coloured serum ; and again upon this, water. The specifically lighter 

 serum is seen interpenetrating in the form of pseudopodial processes into the heavier 

 subjacent salt solution. 



Fig. 2. — Flat diffusion cell filled in below with coloured 6 per cent. NaCl solution. Upon 

 this is superposed uncoloured serum. The subjacent coloured salt solution is inter- 

 penetrating into the supernatant coloured serum, and the uncoloured serum into the 

 subjacent coloured salt solution. 



Fig. 3. — Flat diffusion cell filled in below with uncoloured 6 per cent. NaCl solution, and 

 above with coloured serum. The ascending rills of uncoloured salt solution here 

 unite to form an ascending stream ; and in like manner the descending streams of 

 coloured serum unite to form a cascade. 



This process of descending and ascending can be beautifully demonstrated 

 in a specially shaped diffusion cell, which again was suggested by Dr. Fleming. 

 The cell in question has an upper and a lower reservoir (somewhat after the 

 fashion of a Kipp's apparatus), connected by a much narrower channel, which 

 zigzags down after the fashion of a flattened out spiral (fig. 3). The apparatus is 

 charged with a 6 per cent, salt solution as far as the neck of the upper chamber, 

 and then this last is filled with coloured serum. In a cell thus charged the 

 serum descends through the salt solution as a coloured waterfall, forming at 



* 



