of an indestructable and self-registering barometer, as iron was likewise mentioned by myself, I trust I may be pardoned for throwing out the following suggestions, although I am not certain that Mr. Taylor commented at all on the plan proposed by me, since "No. 2" merely is mentioned, without names being specified. The approval which that scientific gentleman bestows on the self-registering apparatus of No. 2, (deeming it very well worthy of attention and likely to be adopted, if the difficulties before stated in the syphon tube can be obviated), as it works on the same principle as that proposed by myself, induces me to hope that something may be yet done to bring the matter to a successful issue. For observatories and fixed stations a syphon tube of glass, or semi-transparent porcelain (as suggested by a friend), will at once remove the objections justly stated by Mr. Taylor, and in my first letter this recommendation was accidentally omitted, for at that time my attention was more particularly turned to the rendering barometers, if possible, indestructable, and more useful to travellers. The drawing annexed, (Pl. 4. fig. 1.) will, I really think, serve to show that an iron syphon tube may be so constructed as to be easily filled and carried about without the chance of air getting either into the longer limb, or into any situation from which it will not be immediately expelled, when properly prepared for action—(vide figure and remarks). By tapping the barometer while horizontal before working it, the air bubbles will rise to the surface near the stop cock; for since the cistern was nearly full of mercury originally, no additional quantity of air can ever enter it, since the stop cock will be made mercury tight: the passage of air through it will be advantageous. There is room also for the mercury to expand in high temperatures, and after a journey, when the instrument is laid horizontally, trifling particles of air, which may adhere to the sides of the cistern, can be readily expelled by heating the screw end of the cistern after opening the stopper. It will be manifest, also, on inspection before placing the barometer vertically, whether much air has entered the instrument, for the mercury to allow of such occupation must necessarily have been forced through the stop cock or screw; but this seems an impossible case, while the air itself can be so much more easily pressed through those places.
From trials with an iron barometer in my possession, I think the long tube should be of one uniform bore throughout, and not larger at the vacuum end, as before recommended by me, unless made by the most careful and expert workman, since the difficulty of boring the latter, and the trouble of expelling air from it, become very great; while the scale also requires more than ordinary attention.
