and the Psychrometer. 439 



On evaporating the liquid very slowly it was seen that the 

 bubbles arose very near the thermometer that was close by, and 

 there was at first a uniform indication by this of "3° lower tem- 

 perature than by the other. As the liquid lost its temperature 

 and became much cooler than the air it was found necessary to 

 force in the air much more rapidly in order to continue the 

 cooling process ; after a time the two thermometers indicated 

 the same temperature and then the one nearer the inlet tube 

 began to read higher than the other and so continued through- 

 out the experiment, reading half a degree higher for some time. 

 The explanation seems very simple ; at first the slowly flowing 

 air only agitated the liquid near its thermometer and hence it 

 caused a lower reading, afterward the air relatively much 

 warmer, flowing more rapidly through the liquid was not en- 

 tirely cooled to its temperature and hence caused a too high 

 reading of the near thermometer. We may conclude then, that, 

 under both slow and rapid air motion, the thermometer near 

 the plate will not be affected very differently from the plate itself. 



Under the second division we seek for answers to the follow- 

 ing questions. 1st. What effect does the thickness and mate- 

 rial of the plate have ? 2d. Are there any appreciable differences 

 in the hygrometric properties of plates? 3d. Is there any dif- 

 ference in the ease of detection of the dew, etc. ? I have not 

 had access to apparatus by which to settle these questions, but 

 they may be partly answered by using .a single instrument. 

 As quoted above from Mr. Strachan, the thickness of the plate 

 would seem the most important source of error. The follow- 

 ing partial answer to the objection is given. It does not seem 

 probable that the inside of the plate would have precisely the 

 temperature of the thermometer, but theory would indicate that 

 there would be a gradual diminution of temperature from the 

 outside of the plate to a point inside the liquid, and this would 

 bo greater the lower the temperature of the air, as compared 

 with that of the dew point. For testing this question we have 

 the following methods. 1st. We may compare results obtained 

 with a small difference between the air and the dew-point, at 

 which time there would be slight diminution of the tempera- 

 ture of the liquid below that of the front of the plate, with 

 results obtained under just the opposite conditions and which 

 ought to indicate the greater diminution. Such comparison 

 shows an appreciable difference, but this is in precisely the 

 opposite direction from that indicated by theory, i. e., with a 

 small difference between the dew-point and the air, the former 

 is lower, as compared with the air, than with a large difference. 



2d. We may lower the temperature rapidly at one time and 

 slowly at another ; if we have the same result in both cases we 

 can assume that the effect of the thickness of the plate is inap- 



