82 



NATURE 



[May 2 2,, 1889 



they all have the same density, the following numbers 

 •represent their masses, Saturn's mass being unity : — 



: Mimas 1/500,000 I Rhea 1/32,000 



^Enceladus ... 1/270,000 | Hyperion ... i/i,8oo,coo 



Tethys 1/75,000 Japetus ... 1/110,000 



Dione 1/85,000 | 



The mass of Saturn's rings has been found 1/620 that of 

 the planet by observations of the rotational movement 

 which it imparts to the major axes or line of apsides of 

 the satellites. 



The masses of the satellites of Uranus and Neptune are 

 not known to any degree of accuracy. The two satellites 

 of Mars have had their masses deduced from photometric 

 measures, but they are so small— about 10 kilometres in 

 diameter, being no larger than the smallest known 

 asteroids — that the numbers found cannot be very exact. 



Masses of some Stars. 



M. Tisserand rightly gives a dissertation, full and clear 

 withal, of this subject. Sir William Herschel was the 

 discoverer of the relative motions of binary stars in 1802. 

 The obvious conclusion from such a discovery was that 

 the laws of gravitation were universal. Truly, it was 

 not logical to make such an assumption, and some 

 objections have been raised, but the 07tus probafidi rests 

 with those who doubt it. In considering the motions of 

 the components of a binary star system, it must be re- 

 membered that they revolve round a common centre of 

 gravity. It is usual, however, to consider the principal 

 stars as fixed, but augmented by the mass of its satellite, 

 the latter having an orbit which is the mean of the two. 

 Knowing the fall of the satellite to its primary in one 

 second, we may calculate what it would be if at the same 

 distance from it that the earth is from the sun. But we 

 know by how much the satellite would fall towards the 

 sun, since it would fall as the earth. Hence the considera- 

 tion of the two falls will give the sum of the masses of the 

 stars in terms of the sun's mass. 



The following is the formula employed : — ■ 



771 + til' 



m and in' are the masses of the two stars ; M that of the 

 sun ; a is the angle, expressed in seconds, which is sub- 

 tended at the earth by the semi-major axis of the satellites 

 orbit ; p is the "annual parallax" of the binary group 

 expressed in seconds ; whilst T is the time in years of one 

 revolution of the satellite. These are the numbers that 

 have been obtained for four groups, the distances of which 

 from the earth are known : — 



Star. Paralla.x. Magnitude. Mas"^es! 



a Centauri ... o" 8d ... i ... 18 



7J Ca'^siopeias ... o"i5 ... 4 ... 8'3 



70 p Ophiuchi ... o"i/ ... 45 ... 2'5 



<7'^ Eradini ... o"22 ... 45 ... i"o 



Sirius aftd its Companion. 



The article concludes with a complete history of the 

 work which suggested the existence of a companion to 

 Sirius. Bessei had determined the proper motion of 

 thirty-six stars by observations of their right ascensions 

 and comparing with Bradley's, but he found that in the 

 case of Sirius the hypothesis of a uniform variation was 

 irreconcilable with them, and suggested that the irregu- 

 larities might be produced by the action of some obscure 

 body. As a proof that obscure bodies exist in the heavens, 

 the case of Tycho Brahe's Nova is quoted, this being a 

 star which suddenly appeared in Cassiopeia in 1572, and 

 then gradually disappeared without change of place. After 

 Bessel's death Peters found that it was possible to account 

 for the irregularities by the supposition that Sirius de- 

 scribed an orbit in fifty years whose eccentricity was 

 about o"8. Safford, in 1861, from a discussion of the 

 declinations of Sirius, came to the same conclusion as 



HJj.T.; 



Peters ; whilst Auwers, in 1862, after investigating about 

 7000 right ascensions and 4000 declinations, found the 

 time of revolution to be forty-nine years, and the eccen- 

 tricity o-6oi. At the same time as Auwers was engaged 

 with his calculations, Alvan Ciark discovered a small 

 star only about 10" from Sirius, which appeared to be the 

 companion. Future considerations supported the surmise, 

 and proved that this body was precisely what was re- 

 quired to account for the orbit of Sirius round the 

 common centre of gravity. 



If Gill's measure of the parallax of Sirius be taken as 

 correct, viz. o"-38, the sum ot the masses of the two stars is 

 equal to 4-4 that of the sun. Sirius has about twice the 

 mass of its companion, and they are separated by a dis- 

 tance a little more than twice the distance of Jranus 

 from the sun. 



From a discussion of similar little irregularities in the 

 proper motion of /? Cassiopeias, Struve found its mass 

 to be 6-6 compared with the sun, whilst its companion 

 was 17 times as great. 



A reflection on the inability of astronomers before 

 Copernicus to make such measurements as those pre- 

 ceding, concludes this retrospect. 



R. A. Gregory. 



A NEW FORM OFREGENERA TIVE GAS-LAMP. 



T7ROM the time when Mr. Frederick Siemens first 

 -»- introduced regenerative gas-burners, now ten years 

 ago, down to the present day, this method of burning gas 

 for illuminating purposes has been adopted all over the 

 world, and has come to the assistance of the gas com- 

 panies by illustrating the fact that, with proper appliances, 

 gas can produce the same brilliant effects as are ordinarily 

 produced by means of electricity, at much less expense 

 both as regards first cost and working. We would explain 

 that in regenerative lamps the heat which is usually 

 wasted in ordinary burners is to a great extent returned 

 to the flame. The manner in which this result is brought 

 about is by intercepting, by means of a regenerator, the heat 

 passing away with the products of combustion, and apply- 

 ing the heat thus saved to raise the temperature of the 

 air which feeds the flame, thus increasing the temperature 

 of the latter, and its illuminating power ; for it may be 

 admitted that the higher the temperature of a body ren- 

 dered incandescent by heat, the greater is the proportion 

 of light rays emitted out of the total amount of energy 

 radiated. 1 his being the case, the amount of heat carried 

 from such a source of illumination to the surrounding 

 atmosphere by conduction and convection must be less 

 than in the case of a burner consuming the same 

 quantity of gas burning at a lower temperature, which 

 circumstance, combined with the well-known economy 

 resulting from the use of these burners, accounts to a 

 great extent for the popularity which regenerative lamps 

 have attained. 



Mr. Frederick Siemens has lately introduced a new 

 form of regenerative gas-lamp, vhich we understand is 

 highly efficient, and is in consequence being largely 

 adopted ; its construction is shown m the accompanying 

 diagram. It is known as the Siemens inverted type, and 

 is produced in various ornamental designs, which have 

 been much admired. Alter passing through the governor 

 A, and the tap b, the gas enters an annular casing ; in 

 the lower portion of this, a number of small tubes are 

 fixed, forming the burner, Jrom which tubes the gas passes 

 out in separate streams. By this means, combustion of a 

 very perfect character takes place, as the air is directed 

 round each separate stream of gas, and thus enabled to 

 combine most intimately with it. Within the circle of 

 small tubes is a trumpet-shaped porcelain tube, d, and 

 around the outside and inside of this the gas burns, 

 downwards and slightly upwards, as indicated by arrows. 



I 



