TRANSACTIONS OP THE SECTIONS. 51 



The average value of c in the experiments is 28-50, while the vahie arrived at by 

 purely mathematical considerations is 28-55. This close correspondence is the more 

 satisfactory -svhen the difficulty of accurately measuring short intervals of time is 

 borne in mind. 



The formula v=A/px28:55 may therefore be adopted as correct. The author 

 has constructed a Table by means of it, showing the velocities which correspond to 

 the various readings of the anemometer; also a Table showing the coii-ection to be 

 made for variations in the temperature of the air whose speed is to be measured. 

 The corrections to be made for small variations in barometric pressure are unim- 

 portant. M'hen it is necessary to make the correction, recoui'se must be had to 

 the formula 



according to the circumstances of the case. In the former the velocity is given in 

 feet per second of air measured under the barometric pressure existing in the air- 

 channel, in the latter it is given in feet per second of air measured imder a 

 pressure of 29-92 inches of mercury. 



It may be asked if allowance should be made for the expansion or contraction 

 which win take place in the ether of the manometer when exposed to varying 

 temperatures. The variations of temperature to which the manometer itself are 

 exposed are not great, being those of the external atmosphere only. It will be 

 foimd that for a variation of 10 degrees the en-or introduced is about one per cent. 

 Thus if the manometer has been exposed to a temperature of 50°, and the speed 

 of the air experimented on is by calcidation 10 feet per second, the real speed will 

 be 10-1 feet. If the temperature of the ether in the manometer was 70°, then the 

 real speed in place of 10 feet -will be 99 feet per second. 



In order to test by experiment the corrections of the formula for making allow- 

 ance for variations in the temperature of the air whose velocity is to be measured, 

 the following trial was made : — In a furnace, constructed for an experimental 

 purpose, a current of air entered through a pipe 9 inches diameter at a temperature 

 of_ 170° ; after traversing channels of red-hot brickwork, it passed out through a 

 C-incli pipe at a temperature of 560°. 



The reading of the anemometer at the inlet flue was 0-012 inch. Eeferring to 

 the Table, the speed given is 3-127 feet per second. The correcting figure for the 

 temperature 170° is 0-9083. Multiplying the two together, we have 2-84 feet per 

 second as the speed of the air measm-ed at 60°. The quantity of air passing was 

 therefore 1-255 cubic feet per second. 



At the outlet pipe the anemometer reading was 0-102 inch, showing by the Table 

 a speed of 9-118 feet per second, to be multiplied by 0-7137, the figure of correction 

 for the temperature 560°. This gives 6-508 feet per second for the speed of air 

 measured at 00° ; therefore the quantity of air passing was 1-278 cubic feet per 

 second. The error is less than two per cent. Such an approximation is perhaps 

 as close as could be expected in measurements of this kind, and may, it is thought, 

 be taken as a confirmation of the general coi-rectness of the formulae. 



Bescrijytion of a New Self-recordincf Aneroid Barometer. By F. Maetin. 



On the Maury Barometer, a new Instrument for Meastiriny Altitudes. 

 By Feederick T. Mott, F.R.G.S. 

 The author said that at present there was no instrument in the hands of the 

 engineer by which he could make a rapid survey across moimtainous or hilly 

 country with ease and accuracy. At the suggestion of Captain Maury, Mr. E. T. 

 Loseby, a well-known chronometer maker, had invented a pocket barometer, which 

 promised to supply this desideratum. 



The Maury barometer, lilie the aneroid barometer, measures the atmospheric pres- 

 sure by the expansion and contraction of a vacuum-box ; but this box is of larger size 

 in proportion to the instrument, and of superior construction ; and the measuring- 

 machinery consists simply of a fine micrometer-screw attached to the index, and a 



4* 



