442 SCIENTIFIC tiECOiRt) EOR. 1884 



internal diameter, closed at one end and open below, is introduced a 

 thread of mercury about 250""" in length. This incloses a certain vol- 

 ume of air, about 40 cubic centimeters. The tube is supported on a 

 stand, which has a scale on white paper divided into centimeters, and can 

 be phxced on the table with either end vertical. When the closed end 

 is uppermost, the pressure on the inclosed air is the difference between 

 the barometric height and 250; and when inverted, it is the sum of the 

 barometric lieight and 250. If the barometer stand at 750"'"', the press- 

 ure in the former case is 500'"'" and in the latter IjOOO"'"". Consequently 

 the volume is reduced from 40 cubic centimeters to 20. {Phil. Mag., 

 July, 1884, XVIII, 80.) 



DiakoHoff has devised a new form of siphon barometer, which, being 

 easily filled, is more readily portable than the ordinary forms. The 

 barometer tube proper, which at its upper third has a diameter of 0.012 

 meter, is coutiuued above by a capillary tube of the same length, which 

 turns downward. At bottom it opens into the side of a larger tube, the 

 upper end of which is closed with a cork, while the lower end is drawn 

 out to enter one end of a rubber tube, the other end of which is fastened 

 to the lower end of a funnel tube the length of the barometer and a 

 little more. After the apparatus has been thoroughly cleaned, it is 

 filled by jiourlDg mercury in the funnel tube, whence it rises through 

 the barometer tube and overflows at top, the excess of mercury falling 

 through the small tube into a reservoir below, into which its end dips. 

 When all the tubes are full, the rubber tube is detached from the funnel 

 tube, and a i)ortion of the mercury is allowed to flow out till the column 

 stands at the proper height. It is read by means of a graduated scale 

 and two verniers. {J. Phys., II, in, 27, January, 1884.) 



Grumuach has suggested the use of a vacuum tube in connection 

 with the barometer, so that by electrical means the rarefaction may be 

 ascertained. He states that the opinion entertained until recently that 

 the electric discharge would not take place through the best attainable 

 vacuum is not absolutelj^ true, since recent experiments have shown 

 that the passage of the electric discharge depends largely upon the form 

 of the electrodes. In a vacuum in which the spark will not pass if the 

 electrodes are simple wires, it readily passes when they are of the form 

 of balls or cylinders. The author's barometer is connected with a mer- 

 cury pump and with a Geissler tube, operated by a small induction coil. 

 As the exhaustion proceeds, the nitrogen bands first disappear, then the 

 lines F and G of hydrogen and some lines of mercury. Finally the 

 I)hosphorescence of the glass appears first at the negative end, then at the 

 positive end, and finally throughout the tube. At last the discharge 

 is totally arrested or is slowly intermittent. Comparisons with this nor- 

 mal barometer show difl:erences too large to be attributed to errors of 

 observation. The author finds that a difference of 0.07° C. produces 

 an error of 0.01'"'" in the barometer reading. ( Wied. Ann., xxi, G98 ; 

 J. Phys., II, III, 264, June, 1884.) 



