Latent Heat of Evaporation of Benzene. 15 



caused by the evaporation had ceased, and we were thus able 

 to determine when the flask and connecting tubes were 

 saturated. The rod was not finally withdrawn until the 

 dropper was introduced : thus the space being saturated, no 

 further evaporation took place ; and we believe that from 

 Experiment X. onwards any error arising from evaporation 

 in the time preceding the experiment may be disregarded. 



After we had completed our experiments, we made some 

 observations with the object of ascertaining the probable 

 magnitude of the error in Experiments I. to VII. due to 

 evaporation when the flask had not been previously saturated. 

 A dropper was placed within the evaporating flask for 20 

 minutes and kept at a temperature of 30° C. It was then 

 withdrawn, and the loss determined. As might have been 

 expected, it varied slightly according to the droppers used, 

 probably owing to the different sizes of the capillary opening. 



In 20 minutes, dropper II. lost 10 milligrams, 



>J 5) JJ 111. ,, 1Z ,, 



TV 13 



After Experiment VII. we adopted the plan of noting the 

 time from the insertion of the dropper to the commencement 

 of the experiment, but unfortunately we had not previously 

 done so. We consider that the average time in these 

 experiments must have been from 10 to 1 5 minutes ; the loss 

 during this time would therefore appear to have been about 

 6 milligrams, or rather more than j^. Thus the values of L 

 resulting from Experiments I. to VII. are probably too low by 

 about - 12. Fortunately we have six independent experi- 

 ments at 30° C. (Xos. XXIX. to XXXIV.) in which this 

 cause of error was absent. 



We have entered into this matter fully, as it shows the 

 importance of extreme attention to details in work of this 

 kind, and also it was necessary to explain why we practically 

 neglected Experiments I. to VII. when drawing our final 

 conclusions. The same cause of error would slightly affect 

 Experiments VIII. and IX. at 40° C. At this time, however, 

 we were attempting to minimize the evil by allowing as short 

 a time as possible to elapse between the introduction of the 

 dropper and the commencement of an experiment. We have 

 also a note of the time, which in both cases was less than 

 4 minutes. The loss during this interval would probably not 

 have affected the resulting values of L by more than *05, and 

 we therefore do not consider it necessary to reject these 

 experiments. 



At the close of an experiment, when observation of the 



