208 Van Name and Bosworth — Rates of Solution of 



metallic contact between disk and holder, the three horizontal 

 projecting arms which grasped the disk at three points on its 

 circumference were coated with a heavy layer, renewed from 

 time to time, of a commercial pyroxylin preparation known as 

 •• New Skin/'' The slight spring of the platinum wire sufficed 

 to hold the disk firmly in place. 



Variations in the diameter of the disks, since they affect 

 the surface area to a relatively large extent, were avoided in 

 the following way. The metals used were carefully rolled 

 out to a uniform thickness of 0'5 mm and cut into pieces 

 slightly larger than desired. A pair of specially prepared pat- 

 terns in the form of thick circular disks of hardened steel were 

 then clamped in concentric position on opposite sides of the 

 metal sheet, and the edges of the latter carefully trimmed and 

 tiled away to exact coincidence with the circumference of th'e 

 patterns. The diameter of the disks so prepared was 38 - 3 mm 

 with a maximum variation of about 0'05 mm . Before using, the 

 disks were always cleaned with hot caustic alkali and then 

 slightly etched with acid to insure uniform action over the 

 whole surface. 



The velocity constant was calculated from the expression 



v c 



K = 2 - 3 — — log — , in which v is the total volume of the 



liquid, and c, aud <? a are the amounts of thiosulphate needed to 

 titrate a 20 ec sample at times t 1 and t^. 



The rate of stirring was the same in all the experiments here 

 recorded, 170 revolutions per minute. Variations in the rate 

 seldom exceeded zfc 1 per cent, and as great care was taken 

 to adjust the mean rate to 170 revolutions, the fluctuations 

 affected chiefly the single values of the constant, and must have 

 had an almost negligible effect upon the final result, i. e., upon 

 the average value of the velocity constant for the given exper- 

 iment. 



In many of the experiments a further precaution was taken. 

 With the aid of a stop-watch, and of a bell connected with 

 the speed indicator so as to ring after every 100 revolutions, it 

 was possible without interfering with the experiment to accu- 

 rately determine the mean rate of stirring for the whole of 

 each separate reaction period. The corresponding values of 

 the velocity constant were then corrected to 170 revolutions on 

 the assumption that the value of the constant is proportional 

 to the 4/5 power of the rate of stirring.* Although this pro- 

 cedure gave greater security against possible variations in the 

 stirring, the effect of the corrections upon the final result 



* This relation was shown to hold approximately by the work of Van 

 Name and Edgar. 



