It will be easily understood that as the bobbin hangs freely in the center of a very strong magnetic field (formed by the four poles of the two permanent magnets), the slightest current sent through the bobbin will cause the bobbin to be attracted from one direction, while it will be repelled from the other, according to the polarity of the current transmitted. As the relay has a very low resistance, it is evident that it will become an acceptable auxiliary in our central office, particularly when used as a "calling off" signal, as by its use the ground deviation, so objectionable and yet so universally used for "calling off" purposes, can be entirely avoided, and the relay left directly in the circuit, as is being done here in Paris. R. G. BROWN. Paris, September 12, 1882. * * * * * THE PLATINUM WATER PYROMETER. By J. C. HOADLEY. The following description of the apparatus used for the determination of high temperatures, up nearly to the melting point of platinum, is offered in answer to several inquiries on the subject: The object to be attained is a convenient and reasonably accurate application of the method of mixtures to the determination of temperatures above the range of mercurial thermometers, say 500° F., up to any point not above the melting point of the most refractory metal available for the purpose, platinum. A first requisite is a cup or vessel of convenient form, capable of holding a suitable quantity of water, say about two pounds avoirdupois. Berthelot decidedly prefers a simple can of platinum, very thin, with a light cover of the same metal, to be fastened on by a bayonet hitch. For strictly laboratory work this may be the best form; but for the hasty manipulation and rough usage of practical boiler testing something more robust, but, if possible, equally sensitive, is required. The vessel I have used is represented in section in the accompanying cut, Fig. 1.