1883.] J. Bridges Lee — Instrument for determining the dew-fall. 67 



(4.) Connect this cylinder at one end rigidly with a solid cylinder of 

 iron of corresponding diameter. 



(5.) Connect the free end of this iron cylinder with a thin solid uni- 

 form rod of iron cased externally with a thin hollow cylinder of Platinum. 

 (6.) Connect the free end of this rod with the lower surface of a thin 

 horizontal sheet of metal one foot square, or of any determinate size such 

 as may be found most convenient in practice. 



(7.) Cover the upper surface of the plate with a moderately thick 

 layer of non-heat-conducting material. 



(8.) Cover the upper surface of this non-heat-conducting material 

 with a thin film of lamp-black or such other material as it shall be found 

 most convenient to employ as a radiator of heat. 



(9.) Close the main hollow cylinder above with a horizontal disc of the 

 same material or of glass with a circular aperture in the centre of suffi- 

 cient diameter to allow of the free passage of the composite rod up and 

 down. 



The essentials of the apparatus will then be complete, except perhaps 

 that it may be found necessary or advisable to interpose between the top 

 of the composite iron platinum rod and the horizontal disk a short cylin- 

 drical rod of ivory to prevent downward conduction. 



The relative proportions and dimensions of all the various materials 

 employed should be such as to secure the following conditions : 



(a.) That the elongated composite cylinder (with its supported plate) 

 should float vertically in the mercury, the plate being carried when dry to 

 a considerable height above the top of the containing cylinder. 



(J).) That a very slight weight on the upper surface of the plate shall 

 cause an appreciable depression of the composite float. 



(c.) That the effects of heat and cold, producing expansion and con- 

 traction of all the various components of the apparatus, may be such that 

 they will compensate each other. 



(^d.) That the upper radiating surface should be fairly constant in 

 character and not liable to be easily abraded by wind or injured by the 

 various other atmospheric changes to which the instrument will be liable 

 to be exposed. 



These results can perhaps all be obtained by the aid of a little arithme- 

 tic and a little initial care in the construction of the instrument, but, if we 

 suppose that all the conditions cannot be completely fulfilled, it will always 

 be possible to furnish each instrument with a table of errors. 



The whole instrument should be provided with a cover, and it should 

 be uncovered only on cloudless nights, when, in proportion to the quantity 

 of dew which accumulates on the upper plate, the float will be forced down- 

 wards into the mercury in which it floats. 



