TRANSACTIONS OF THE SECTIONS. 25 



Thve^iKales of temperature, ranging from 25° to 100°, and tlius exceeding the 

 range oithe temperature of sea-water, are fixed upon the upper end of the plate, to 

 the lower part of which the bars are attached. 



Upon one of these scales the present temperature is shown by the needle or 

 pointer (E), regulated by the motion of the bars. Two other needles moving upon 

 the same pivot are useful as indexes, the needle (c/) being pushed by a pin on the 

 pointer (E) to the highest temperature attained, and the needle (F) to the lowest, 

 where they are retained in position by still' friction on their axis. 



It wiU be observed that lateral concussion is avoided by the employment of two 

 bars attached to the needle at points equally distant from its centre. 



The experiments made by Sir F. Leopold M'^Clintock with the deep-sea pressure- 

 gauge showed that further improvements in the instrument were essential. 



The pressm-e-gauge used by him in 1860 consisted of a brass cylinder, with a 

 solid piston rod, in which the water was compressed in descending by the piston 

 rod, forced inwards by the pressure of the denser external water imtil it became 

 equally dense. 



In filling this metallic cylinder, .it was found that so many air- bubbles adhered to 

 its internal surface as to interfere very materially with the results of experiment. 



To obviate this difficulty, a cylindrical vessel of glass was adopted, with a gra- 

 duated scale upon a long neck or stem, and closed -with an elastic plug of caout- 

 chouc, which when descending was forced inwards by external pressure, compressing 

 the water in the pressure-gauge, and pushing along the gi-aduated scale a spring 

 index, which remamed and marlied the degree of compression, when the elastic 

 plug was forced back again to the top of the scale by the expansion of the com- 

 pressed water, as the pressm-e-gauge was raised again to the siuface. 



In this instrument the absence of air-bubbles was easily secured, but the difficulty 

 of regulating with exactness the lubrication of this plug, and thus regulating the 

 friction against the inner surface of the stem, rendered the results of experiments 

 still somewhat uncertain. 



The form of instrument now used combines the advantages of a transparent glass 

 vessel with the absence of friction. It consists of a cylindrical glass vessel (A), 

 having at one end a sphere (B) perforated with one or more small apertures, and 

 covered with a close-fitting elastic diaphragm, and at the other end an oiled silk 

 valve (D), resembling the valve of an air-pump, and admitting water into the 

 cylinder under external pressure when descending, until the water in the pressure- 

 gauge is of equal density with the external water. Upon a shoulder projecting 

 near to the upper end of the cylinder a small glass vessel (H) is fixed, covering the 

 sphere (B) ; and from it rises a small tube (I), graduated on such a scale that each 

 part or degree is eqiial to , „'„ „ part of the volume of the water contained in the 

 cylinder, and consequently one-tenth of a degree is equal to jt.Vto part. 



To protect the valve from gi-it, and also to prevent the admission ot air into the 

 cylinder, an elastic waterproof bag (E), with a perforated metal cover, is filled 

 with water, and screwed on to the valve under vxiter to exclude air ; and the valve 

 is then screwed into the end of the cylinder, which should be quite full of water, 

 the superfluous water escaping through a slot in the valve-screw, so that air is ex- 

 cluded from the cylinder. 



When the pressure-gauge is thus filled, and the case fastened, it is ready for 

 sinking valve downwards, on a sounding line, into water of gi-eat depth. 



AVhen the pressm-e-gauge sent down is again raised into water of less density, 

 the increased quantity of water forced into it by external pressure is retained ty 

 the valve in consequence of the elastic diaphragm covering the sphere yielding 

 readily to its expansion, and being distended in proportion to the quantity of water 

 forced in ; and an equal quantity of water is expelled, through the open tube, from 

 the small glass vessel (II ) covering the sphere. 



This displacement of water from the vessel (H) is the measure of the depth to 

 which the pressure-gauge has been lowered, and the amount of it is ascertained 

 readily by adjusting the water in the tube (I) to the degree 2000, and then turning 

 the valve-screw (D) very slightly, to allow the excess of water to escape through the 

 slot in the valve-screw, and the elastic diaphragm to sink again closely on the sphere, 

 and then reading off on the scale-t,ube the lieight of the water remaining in it. 



