the Elementary Laiv of Hydrodiffusion. 497 



the temperature of the room regularly rose during the forenoon, 

 till about twelve o'clock, about 2° C, to fall exactly as much, 

 extremely slowly and regularly, from that time till the next 

 morning. The observations, which in each series Avere carried 

 on throughout both day and night, at intervals of three hours 

 each, showed that the diffusion proceeded with the utmost 

 regularity during the period from about 1 p.m. to about 7 A.M. ; 

 while for the six hours of the forenoon they gave, day for day, 

 a somewhat devious course. This indicated that variations of 

 temperature even to the amount of only about 2° considerably 

 disturb the course of diffusion, and that pure results can only 

 be obtained at as nearly as possible constant temperature of 

 the diffusion- vessel. (From the experiments afterwards com- 

 municated it will be seen that the method of observation I 

 have employed is so sensitive that it makes most distinctly 

 evident the influence of a temperature-variation of as little as 

 J°.) Professor Wolf had the kindness to place at my disposal 

 for a long time the cellar-rooms of the Federal Observatory. 

 I there carried out two very protracted series of observations. 

 Unfortunately, in the middle of the first series a sudden and 

 very considerable fall of the outside temperature occurred, the 

 influence of which, although much weakened, extended down 

 even into the cellars. Hence I forbear to communicate the 

 results of that series of observations. During the second 

 series the temperature varied extremely little and very slowly, 

 as may be seen from the readings given in the following Table 

 (p. 499) ; for this second series the experiment was arranged 

 as follows. 



On the 9th of November a solution of zinc sulphate, freed 

 as much as possible from air, and possessing the concentration 

 z 2 — 0*31 82 (that is, a cubic centim. of solution contained 

 0*3182 gram of anhydrous salt), was brought upon the lower 

 amalgamated zinc electrode up to the height / 2 = 1*98 centim. 

 At 7 a.m. to the moment, over this solution a second, with 

 the lower concentration z x — 0*1520, was superposed, to the 

 height Z 1 = l , 01 centim., so that the surface of separation 

 between the two solutions proved perfectly even and reflecting. 

 Then the second zinc electrode was lowered, by means of a 

 suitable guide, upon the upper bounding surface of this second 

 layer till it came into contact with the solution, and was fixed 

 in that position. The diffusion-vessel having been closed air- 

 tight by means of paper steeped in bees'-wax and some addi- 

 tional bees'-wax, it remained untouched upon a pillar of the 

 room during the whole twelve days of the observation. The 

 first reading was made immediately after the preparation of 

 the experiment : the further readings followed six times daily, 



