122 Description of Sketches of a Sdf-Registering Barometer. [Jan-, 



The wire, 1900 feet in length, at 9 annas per 100 feet (the retail 

 Madras price), would cost 10 rupees 11 annas. The planking about 25 

 rupees, and the iron, workmanship, &c. about 25 more —total about 60 

 rupees. 



The perusal of the foregoing memorandum some time back led me 

 to think, that iron wire might most advantageously be applied in India 

 (where we are ruined by white ants) as a substitute for woodwork in 

 roofing, and the various calculations I have made convince me, that the 

 day is not very distant, when this valuable material will be extensively 

 employed for that and other purposes. I shall venture to forward to 

 you at a future opportunity, some remarks on roofs composed of hollow- 

 cylinders, and my calculations respecting those formed of wire instead 

 of wooden rafters and reaper s, 



P. S. — It has occurred to me, that, should iron wire be found to cor- 

 rode rapidly on the sea coast, and not be effectually preserved by paint, 

 the elastic gum of the banyan tree, or other usual compositions, that coir 

 rope, which is preserved by sea water, might prove sufficiently perma- 

 nent for the purposes above alluded to. 



■VIII. — Description of Sketches of a Self-Registering Barometer on the 

 Float principle. — By William Gilchrist, Esq. of the Madras Medi- 

 cal Establishment. 



The subject of a Self-Registering barometer having ccupied a part 

 of several numbers of this Journal, it is with considerable hesitation I 

 venture to make further allusion thereto, and certainly would not do so, 

 did it not appear to me, that the accompanying sketches reduce th e 

 instrument to the greatest degree of simplicity of which the principle 

 on which it is based admits. 



In applying the float principle to practice, it will easily be compre- 

 hended, either that the tube may be made moveable and the cistern fix- 

 ed, or vice versa^Fig. \ ( Plate) represents the former of these varieties, 

 fig. 2d the latter. 



In the experimental barometer alluded to in my last communication^ 

 the float buoyed up not only the tube, but also the ballast required to 

 keep the tube perpendicular. This plan requires a considerable quan* 



