544 lioyal Society. 



ration, as well as the great length of the mains ; and he therefore 

 Avould expect the distribution of coal-gas in cities to exemplify ap- 

 proximately the laws of gaseous transpiration. The velocity of coal- 

 gas should be \'515, that of air being 1, under the same pressure. 

 And with a constant propulsive pressure in the gasometer, the flow 

 of gas should increase in volume with a rise of the barometer or with 

 a fall in temperature, directly in proportion to the increase of its 

 density from either of these causes. 



These laws, it will be observed, are entirely different from those 

 which direct the passage of gases through an aperture in a thin plate, 

 or their flow into a vacuum as it is usually said, and could not be 

 deduced, like the latter, from our speculative ideas respecting the 

 elastic fluids. 



" On the Automatic Registration of Magnetometers and Me- 

 teorological Instruments by Photography." — No. III. By Charles 

 Brooke, M.B., F.R.S. 



The author describes the construction of an apparatus for regis- 

 tering the variation of the thermometer and psychrometer on one 

 sheet of paper. As in the apparatus for registering the vertical force 

 magnetometer, described in a former paper, the photographic paper 

 is placed between two concentric cylinders, placed with the axis ver- 

 tical, and carried round on a revolving plate or turn-table by the 

 hour-hand of a time-piece, which makes half a revolution in twenty- 

 four hours ; thus each half of the paper presents a record of the 

 variation of one instrument during twenty-four hours. The scales 

 of the instruments are continuously impressed on the paper by placing 

 fine wires opposite each degree across the aperture through which 

 the light falls on the stem ; the light transmitted by the empty bore 

 is intercepted by these wires, and the darkened portion of the paper 

 is marked by a series of parallel pale lines corresponding to each 

 degree : thus the distortion of the scale arising from the varying di- 

 rection of the pencils of light is corrected. Every tenth degree is 

 marked by a coarser wire, and therefore a broader line, as also the 

 points 32°, 54<°, 76°, 98° ; one at least of these points will occur on 

 each register, and the position of the extra broad line serves to iden- 

 tify the part of the scale to which the register relates. 



An alteration in the mode of adjusting the wick of the camphine 

 lamps described in a former paper is mentioned, by which the chance 

 of smoking is considerably diminished ; likewise the successful ap- 

 plication of naphthalised gas, and of an oil-lamp, to photographic 

 registration. 



The paper concludes with the description of a new method of de- 

 termining the scale and temperature coeflScients of the force magne- 

 tometers, by which a greater degree of accuracy is presumed to be 

 attained than by the methods ordinarily employed. Two magnets 

 designed for self-registering instruments for the observatories at 

 Cambridge and Toronto, having been submitted to this method, gave 

 consistent results which indicate the law of the temperature coefficient 

 to be sensibly different from that which has hitherto been assumed. 



