certain Dilute Aqueous Solutions. 387 



and those for potassium carbonate from Gerlach's values as 

 given in Landolt and Bornstein's Physikalisch-Chemische 

 Tahellen. The densities for water were taken from Marek's 

 values*. 



Method of Observation. 



Before each series of readings the tray, the glass plate, the 

 brass hoop for cleaning "the surface of the liquid, the ther- 

 mometer, and the stirring rod were washed, first with tap- 

 water, then with an alcoholic solution of caustic potash, then 

 with running tap-water, then with chromic acid, and, finally, 

 they were well rinsed in running tap-water. Then everything 

 was placed in position in the iron box, water was run into 

 the tray until the lower edge of the glass plate dipped just 

 under the surface. I then expanded the hoop, put the lid on 

 the box, and brought the iron slab to rest as soon as possible. 

 Then the forks were set in vibration, the spider-line of the 

 telescope was brought into coincidence with a crest near the 

 middle of the pan, and the amplitude of the vibrations was 

 increased until the crest was almost out of focus. As stated 

 before, the waves appear stationary. Then the shunt circuit 

 by which the driven fork (F 2 ) was run was broken and the 

 fork was allowed to come to rest at its free period. The 

 driving-fork (F x ) being still vibrating, the crest will have an 

 apparent motion unless the free period of the driven fork 

 coincides with the running period of the driving one. If the 

 two did not coincide, the weights on the driven fork were 

 moved until they did. By care the depth of the water can 

 be adjusted so that this correction is seldom necessary ; but 

 unless the two forks are in unison the crests are almost sure 

 to oscillate, on account of the imperfection of the mercury 

 contact. The plate which generates the waves should never 

 leave the liquid, but it should dip below the surface as little 

 as possible. If it dips deeply into the liquid the waves always 

 appear distorted. 



Having brought the forks to unison, the spider-line is 

 brought successively into coincidence with three consecutive 

 crests near scale-division 17. Then the carriage is screwed 

 along and readings are taken on the fortieth, forty-first, and 

 forty-second crest beyond the one on which the first setting 

 was made. The thermometer is then read, the hoop is lifted 

 from the water, contracted, and replaced in the water, and 

 expanded, and another series of six readings is taken. Then 

 a known weight of salt is added, and the solution is stirred 

 until the salt is dissolved. Then readings are taken as for 



* Wied. Ann. xliv. p. 171 (1891). 



2F 2 



