At low water, the pressure of the atmosphere balances a column of mercury rising in a glass tube to a height proportionate to such pressure. In measure as the level of the water rises, the pressure on the mercury in the receptacle increases, and causes the metal to rise in the tube. The higher the level of the sea, the less becomes the sum of the resistances of the rheostat, since the column of mercury puts in short circuit all the divisions of the rheostat, whose contacts are comprised in the height of the column. From these variations in the resistance of the circuit naturally result variations in the current from the battery, B, at the registering station. To the variations in intensity of the current in the circuit there correspond variations in the attraction of the solenoid for the core that transmits these motions to the balance that carries the registering style, which latter amplifies or reduces them. The same transmitter suffices for various registering stations arranged in series, as shown in Fig. 2. The variations in the resistance of the circuit, due to variations in the temperature, and the variations in the height of the column of mercury, due to atmospheric variations, etc., are, according to the inventor, of no importance. It would evidently be possible, on the same principle, to construct an apparatus for registering the indications of a thermometer at a distance. Such is the principle of Mr. Gime's apparatus. We do not believe that they are entirely closed to criticism. What, in fact, are the conditions essential for their proper working? Evidently: (1) the constancy of the battery used; (2) a rigorously accurate adjustment. This latter condition, is easily realized; but the same is not the case with the former. Of what elements shall this constant battery be formed? Mr. Gime recommends the use of the Latimer-Clark elements. Every one knows that the Latimer-Clark element is now the best standard of electromotive force; but let us not forget that this is on condition of its being employed in open circuit. Now, it is not a question here of an open circuit, nor even of infinitely weak currents, since in the line we have a solenoid whose core must set in motion a whole system of connected pieces. We do not see any possibility of employing Latimer-Clark elements; on the contrary, it seems to us indispensable to select piles of large discharge, since the solenoid, S, will attract nothing at all unless a notable quantity of energy is expended in it. Is there a pile of this kind so constant as not to render a rigorously accurate adjustment illusory? Therein lies the entire question, and for our part we hesitate to pronounce ourselves in the negative.--_La Lumiere Electrique._ * * * * * A CLINICAL LESSON AT "LA SALPETRIERE." [Illustration: THE SALON OF 1887.--A LECTURE IN THE DISPENSARY AT LA SALPETRIERE.--Painted by M. Andre Brouillet.--M. Dochy. Engraver.]