162 Remarks on Self -registering Thermometers. 



lance the use of a magnet to adjust Six's indices, which often 

 requires to be a powerful one, and to meteorologists who ob- 

 serve a dew-point hygrometer it will be no inconvenience 

 whatever. The method proposed by Lord C. Cavendish to 

 adjust his minimum thermometer appears to be precluded in 

 practice, where the tubes are of moderate bore, as it sup- 

 poses the free passage of the mercury in drops through the 

 alcohol in the tube. Perhaps in executing the thermometer 

 I have now proposed, it might be advisable to have a detach- 

 ed thermometer for the positive temperatures at the moment 

 of observation, and this would preclude the necessity of hav- 

 ing any fluid but mercury in the upper portion of the instru- 

 ment. The less contact we have between the alcohol and mer- 

 cury, I am inclined to think, the instrument would be more 

 perfect, since the successive passage of two such fluids through 

 the same tube must render it liable to be soiled. Indices, 

 however, at least when they are furnished with springs and 

 moved with the magnet, are, I think, the most detrimental to 

 the perfection and general adoption of the register thermo- 

 meters. Not merely are they troublesome to adjust, and liable 

 to go out of order, but their formation is always imperfect : 

 for it is difficult to extirpate the air from the interior ; and the 

 bulbs and tubes must be so large for the admission of the in- 

 dices as to destroy all confidence in their sensibility. The 

 principal advantage I therefore hold out, in the adoption of 

 thermometers similar in construction to the one I have here 

 described is, that they may be made to any degree of delica- 

 cy ; and the finest capillary tubes with small bulbs are in fact 

 more suited to the principles of the instrument, than the larg- 

 est and widest, which can be said of no other species of self- 

 registering thermometers. I need only mention how unfit 

 either Six's or Rutherford's thermometer, as made by the best 

 makers, are for nice experiments. The former has two contacts 

 of alcohol and mercury, as in my thermometer, and two in- 

 dices besides. For ascertaining maximum and minimum tem- 

 peratures for a short period of time, and with any delicacy, 

 such as in sending instruments, by means of small balloons, 

 to the higher regions of the atmosphere, all the ordinary ones 

 «ire quite unsuited ; and it is not till register thermometers 



