62 Mr John Adie on Derv-point Instruments. 



observed ; the space between the outer and inner bulb is near- 

 ly filled with any Hquid which will not freeze by the depression 

 of temperature required for obtaining a dew-point, as alcohol, 

 or water mixed with a quantity of salt : when the instrument 

 is used, the liquid is kept in motion by the hand. 



With this instrument, which is shown in Plate II. Fig. 1, 

 I have obtained constant results, and a dew-point always the 

 same as given by Saussure's method, the greatest difference be- 

 ing half a degree, and that only in three or four instances, — a 

 diff*erence that may very well arise from errors in observation. 

 The temperature of the atmosphere is first found with it as 

 with the common thermometer ; the result is easily obtained, 

 and without the use of much ether in cooling : and the instru- 

 ment is not larger than the common pocket thermometer. 



I may here mention that I had constructed an instrument 

 with a brass covering bulb, into which was inserted a piece of 

 black enamel, on a thin plate of gold, which was in contact 

 with the enclosed liquid ; with this I obtained rather an un- 

 locked for result. When the ether was applied, the brass, being 

 exposed to a great depression of temperature, conducted the 

 heat from the enamel with greater rapidity than the enclosed 

 liquid could impart its temperature ; thus the surface of the 

 enamel acquired a lower temperature than the enclosed ther- 

 mometer, and gave a dew-point higher than the truth. As 

 this effect might also have arisen from the tendency of the sur- 

 faces of different substances to acquire moisture from the at- 

 mosphere, when reduced to a temperature near the point of 

 saturation, to determine this, I made the following experi- 

 ments : — A vessel of brass having a piece of gold, silver, enamel, 

 and glass, set into it, was used to cool water by the slow pro- 

 cess of Saussure, and a deposition of moisture was observed 

 on all the surfaces at the same temperature. The glass took a 

 little longer time, but required no farther depression of tem- 

 perature. An instrument was constructed with a small plate of 

 gold, which gave the same results as when the enamel was used. 



Mr Daniel's instrument is always higher than the truth. In 

 some cases the error amounts to 6 or 7 degrees, as shown by 

 the following table; and the mean of 28 observations gives -{-2.9°. 

 I have also found the indications of this instrument to depend 

 very much on the purity of the enclosed ether used in its con- 



