48 REPOET — 1 887. 



liminary experiments made by him, and the Committee are fully convinced 

 that, if provided with funds, they will be able during the next year to 

 complete experiments which will lead to recommendations, which, if 

 adopted, will place the question of authorised standards on a satisfactory 

 footing. 



Third Report of the Committee, consisting of Pj-ofessors Eamsay, 

 TiLDEN, Marshall, and W. L. Goodwin (Secretary), appointed 

 for the purpose of investigating certain Physical Constants of 

 Solution, especially the Expansion of Saline Solutions. 



The experiments on invaporation described in our last report have been 

 continued, and new series have been begun. The process of invaporation 

 ia so slow that our report each year must necessarily be imperfect in many 

 points. Of all the experiments set up (and some of these have been going 

 on for nearly two years) only four are completed. The method formerly 

 employed, of sealing in glass tubes and opening by breaking the tubes, 

 was found inconvenient, several interesting experiments having been 

 spoiled by various accidents in sealing and opening. We next tried the 

 so-called ' Gem ' and 'Crown ' jars, used a great deal in America for seal- 

 ino- up fruits, &c. They are closed by means of a glass cap, the rim of 

 which is pressed down upon a flange on the neck of the bottle by means 

 of a metallic screw-i'ing working upon a thread below tbe flange. A ring 

 of caoutchouc lies upon the flange and is pressed upon by the rim of the 

 cap. Out of a dozen carefully selected specimens of these jars only two 

 were found to be even approximately tight when tested by the invapora- 

 tion process. For example, experiment 62 shows a loss of 1 •2661 gram 

 of water in 133 days. Still, the jars answer their original purpose, and 

 can be closed tightly enough to prevent the entrance of putrefactive and 

 fermentative germs. The rate of invaporation was found to be very much 

 slower when the jars were used, although larger tubes were employed, 

 and thus larger invaporating surfaces secured. A comparison of A. 2 and 

 C.2 brings out this diff'erence in rates. In A.2 half the quantity of 

 sodium chloride invaporates in 111 days six times as much water as is 

 invaporated by the sodium chloride of C.2 in 133 days. Of course, this 

 is to be explained, in part at least, by the lower vapour tension due to the 

 escape of moisture from the imperfectly closed jar. But, that there is 

 another cause seems to follow from the case of C.3. The jar in this case 

 was almost air-tight, allowing the escape of only 0-2.592 gram of water 

 in 131 days. The total water invaporated during this period was 0'8689 

 gram, as compared with 1'435 gram invaporated in -56 days by smaller 

 quantities of salts inclosed in a sealed tube (see A.l). The jars were 

 very much larger than the sealed tubes, and doubtless this circumstance 

 retarded invaporation when the jars were used. The jars were rejected 

 in favour of wide-mouthed stoppered bottles, the stoppers of which were 

 greased with lard. The loss of water from these bottles was found to be 

 scarcely appreciable. 



In the following tables Ma = mass in grains. Mo = relative numbers 

 of molecules, P = period of invaporation in days, reckoning from the 

 beginning of the experiment, Q = masses of water in the small tubes, 

 R = relative numbers of molecules of water in tlie small tubes. 



