1837.] Capt. Underwood's Plan for an Indeslructable Barometer. 69 



Now, presuming the instrument-maker has thoroughly expelled the 

 air on first filling the instrument, all that the possessor would require 

 to do in order to satisfy himself as to the pressure of air in any portion 

 of the tube, would be to incline it in order to fill the long limb, turn 

 the stop-cock k and screw home the cistern screw. If the mercury 

 now stands at the height mentioned on the instrument, it of course 

 must be as perfect as when it left the maker. The long limb ought to 

 be sufficiently capacious to contain all the mercury in the glass limb, 

 and part of that in the cistern, so that when the stop-cock above the 

 latter is closed, there will be as much air in the cistern as will allow 

 space for the expansion of the quicksilver in the whole tube, while this 

 air is prevented entering the long limb, as the end of this will always 

 be immersed in mercury. A simpler form than the above would con- 

 sist in adopting that of the Englefield barometer, with the addition of 

 a short glass limb rising out of the cistern, the long limb having ex- 

 panded top, to contain the mercury in the short one, and part of that 

 in the cistern as above mentioned. If this ariangement were applied 

 to M. Gay Lussac's barometer, the possibility of air entering the long 

 limb would be much more likely to be prevented than by the inverted 

 cone, which after all, does not prevent the entrance of air, but merely 

 confines what may have got into the lower part of the mercurial 

 column. The instrument also might be carried in any position, which 

 at present it does not admit of without the danger of air entering. 



VI.— -Further Remarks regarding a Plan for an Indestructable Baro- 

 meter. — By Captain George Underwood, Madras Engineers. 



. In the remarks of the Astronomer of Madras, respecting the Inde- 

 structable Barometer and Self-Registering Apparatus, recommended 

 by Dr. Gilchrist, I observe that Mr. Taylor admits that an iron tube 

 can be rendered air-tight by the application of caoutchouc varnish, but 

 adds that this precaution was eventually of no avail on the application 

 of fire, for the purpose of boiling the quicksilver in the tube. Now y 

 since the mercury may be boiled in the tube before applying the var- 

 nish, I cannot understand why the iron should be condemned for the 

 reason alluded to. True it is, that, with such a material, air bubbles 

 cannot be discerned, and with any opaque substance the liability to 

 error is vastly greater on that account, if some means cannot be adopt- 

 ed for ascertaining the existence of so much air as shall vitiate the ob- 

 servations. However, in furtherance of some hasty and brief recom- 

 mendations which I ventured to forward to the Journal on the subject 



