C. Barus — Subsidence of fine Particles in Liquids. 125 



of capillary attraction, "000005 cm , being at least 100 times the 

 molecular radius H,0, it appears that the striking effect of the 

 molecule HC1 in accelerating subsidence, is not an abnormal 

 occurrence, at least from a physical point of view. Other rela- 

 tions are adduced in the Bulletin. 



3. To account for these phenomena as a whole, Mr. Durham, 

 in his second paper, proposes an hypothesis in virtue of which 

 the scope of the action of affinity is enlarged, and suspension 

 regarded as the lower limit of solution. This view is rapidly 

 gaining ground ; nevertheless without concise experimental ref- 

 erence to the density and size of the solid particles and the vis- 

 cosity of the liquid, Mr. Durham's explanation contains no 

 sufficient reason for the observed suspension, nor for changes 

 of rate of subsidence. Prof. Brewer's ingenious hypothesis of 

 colloidal hydrates so constituted that the particles may ulti- 

 mately swell up and float something like gelatinous silica, or 

 even like starch grains, is more direct ; and before further 

 reasons of the cause of suspension are sought the validity of 

 this inference must be tested. This test is feasible, I think. 

 If the phenomenal difference of rate with which the same par- 

 ticles subside in water and in ether is due to volume changes 

 of the particles, then a marked difference in the density of the 

 sediment (clay or tripoli) in water and in ether respectively, 

 must be experimentally demonstrable. Cf. equation (L). I 

 made these experiments with both solids, using two nearly 

 identical density flasks, one for water and the other for ether, 

 in the usual way. The powders were sampled and dried at 

 200° in an air bath for half a day, transferred to the pycnome- 

 ters, dried and weighed when cold. They were' then left in a 

 dessicator for 18 hours and again weighed ; and finally dried in 

 vacuo and weighed. The results throughout were satisfactorily 

 constant. By aid of an apparatus specially devised for the pur- 

 pose, the two flasks were once more thoroughly exhausted 

 (mercury air pump), and then filled with water and with ether 

 respectively, in vacuo, over sulphuric acid. In both cases (the 

 experiments were made consecutively), the vacuum ebullition 

 was kept up for some time to give full assurance of the expul- 

 sion of air, etc. The density of the ether had been previously 

 determined by the same flasks, once for each experiment. 



Thus I obtained for tripoli, 



In water, density == J„=2 , 672, 

 In ether, density =z/ e = 2*697. 



Again, 1 obtained for white bole, 



In water, A w = 2*639, 

 In ether, ^ e z=2*663. 



The manipulation being somewhat difficult, the observed 



