July 12, 1894 J 



NATURE 



249 



inch. If the instrument be tilted through an angle of 

 one second about a line parallel to the plane of the sus- 

 pending wire the mirror turns through an angle of 49', 

 and the magnifying power of the instrument is therefore 

 very nearly 3000. A scale is placed at a distance of ten 

 feet, and 3'44 inches of this scale correspond to a 

 tilt of one second. If the lamp be displaced by ,^ of 

 an inch, the movement can be clearly perceived. It 

 follows, therefore, that it is possible to observe with this 

 pendulum a tilt of less than T.lg of a second, an angle 

 less than that subtended by a line an inch long placed at 

 a distance of a thousand miles. 



It would seem, then, that the bifilar pendulum is 

 admirably adapted for measuring the minute changes of 

 level, which are perhaps the cause of some of the 

 uneliminaled errors in many astronomical and physical 

 inquiries. Indeed, it may be hoped that the time is not 

 far distant when a pendulum of this kind will be regarded 

 as a part of the ordinary equipment of every great obser- 

 vatory. .Again, the bendings of the earth's crust by 

 changes of barometric pressure,' by the ebb and flow of 

 the tides, &c., may be studied, as well as the long-period 

 pulsations produced by violent earthquakes in almost any 

 part of the world. It is not too much to expect, also, 

 that in time we may be able to trace out and measure 

 the slow secular movements of the earth's crust which, 

 after the lapse of ages, become perceptible to the 

 geologist ; and that the vexed question of the origin of 

 lake-basins may receive an answer that will remove this 

 most debatable of subjects from the domain of contro- 

 versy for ever. C. Davison. 



THE SPECTRUM OF OXYGEN IN HIGH 

 TEMPERA TURES:- 



A T a previous meeting, I brought before the Academy 

 ■^~*- a method, founded on the use of electricity, of bring- 

 ing gases under pressure to a high temperature without 

 heating sensibly the recipients which contain them. 



Before rendering an account of the experiments 

 already made on oxygen by means of this method, I 

 shall mention first those which have preceded them, 

 and in which temperatures not exceeding 300 have been 

 realised by means of a line of gas jets playing directly 

 on the tube containing the oxygen. 



The arrangement was as follows : — .K steel tube ten 

 metres long, hned inside with red copper, and closed 

 at its extremities by glass, according to our ordinary 

 modes of closing, was placed in a trough containing a 

 sand-bath. This trough was immediately warmed by a 

 line of a hundred gas jets. The temperature of the tube 

 was taken by means of thermometers metallically con- 

 nected to the tube. 



After having introduced the oxygen at the required 

 pressure, and before the heating of the tube has begun, 

 a good spectrum of a luminous light source is obtained, 

 the beam being thrown along the tube in such a way 

 that any change in the spectrum brought about by heat- 

 ing the gas is perceptible. 



When the jets are lighted, the spectrum changes in pro- 

 portion, as the temperature increases at the same time 

 as the pressure. If the experiment is well conducted, the 

 the pressure of the gas at the end, that is to say, when 

 lights have been extinguished and the temperature has 

 become what it was at first, returns to its original value. 

 To obtain this result there must be no loss of gas during 

 the experiment. 



This loss of gas is caused principally by the lengthen- 

 ing of the pins that unite the pieces of steel which hold 

 the glasses at the extremities of the tube. In order 



• Such ohscrv.-itions would gather addilional interest if made in the neigh- 

 bourhood of a great line of fault. 



; A communication made by Dr. Janssen to the Paris .Academy cf 

 Sciences. Translated from Ld Nature, 



to prevent this lengthening, bands of brass have been 

 placed between the heads of the pins and the disk, the 

 length of which has been calculated to compensate by 

 their expansion that of the pins. Thus the same degree 

 of pressure is obtained at all temperatures. 



The experiments have been made with varied pressures 

 of oxygen. They show that from the ordinary tempera- 

 ture up to about 300" the bands and lines of the spectrum 

 of absorption of oxygen do not undergo modification. 



But quite a new feature is produced. A remarkable point 

 is the very remarkable augmentation of transparency 

 of the gaseous column with the increase of heat, a trans- 

 parency revealed by a considerable augmentation of the 

 brightness and the limits of the spectrum, above all on 

 the side of the red. I shall have to return to the 

 theoretical consequences of this important fact. 



To ascend further in the scale of temperatures, it 

 is necessary to make use of the platinum spiral tube 

 rendered incandescent by the passage of the current. 



I will not repeat the general arrangement of the ex- 

 periment already described. The incandescence of the 



Kic. I. — Experiments with a vertical tube and incandescent spiral. 



spiral is more difficult to obtain if the pressure of the gas 

 is greater. 



The temperature to which the spiral is carried can be 

 determined in different ways: — (i) The thermo-electric 

 couple ; (2) the observation of the increase of the pres- 

 sure of the gas caused by the passage of the current ; (3) 

 and finally, the brightness and length of the spectrum 

 given by the incandescent spiral, when it furnishes the 

 light alone. 



The experiment works thus : — 



The tube being placed in a vertical position, the 

 lamp, which must produce the beam to be analysed after 

 its passage in the tube, is first regulated, and then the 

 spectrum analysing apparatus is arranged. Pressure is 

 then put on, and the constitution of the spectrum having 

 been well observed, a current is made to pass, the power 

 of which is proportional to the temperature which has to 

 be obtained. 



NO. 1289, VOL. 50] 



