September, igo6.j 



KNOWLEDGE & SCIENTIFIC NEWS. 



539 



centre, but are upon one of the minor axes, and on that 

 axis are three times as far from one extremity as we 

 are from the other. Mr. Lewis estimates the longest 

 diameter of the egg as roughly equivalent to six hun- 

 dred light-years, the smallest diameter to three hundred 

 lig-ht-years. The average proper motion for pairs 

 showing- relative motion is three times that of pairs 

 showing no relative motion. 



This important result is far from being the only one 

 vvhieh Mr. Lewis has deduced from his inquiry. fhcrv 

 was a time when the observation of double stars would 

 have been considered suitable enough for amateur astro- 

 nomers, but quite beneath the dignity of a great 

 Government Observatorj ; indeed, it is doubtful whether 

 the earlier Astronomers-Royal would have consented to 

 give the work a place upon the Greenwich programme, 

 seeing that the original purpose of the Observatory 

 was the determination of the exact positions of the 

 stars and the following of the movements of the sun. 

 moon, and planets. Vet even from the stan.dpoint of 

 the severest interpretation of the original warrant for 

 the Royal Observatory, double star work in Mr. Lewis's 

 hands has justified itself, since in not a few cases it has 

 thrown light upon the observation of stars in the transit 

 circle, interpreting and clearing away apparent dis- 

 cordances. On the other hand, in the case of some 

 binaries, the transit circle observations have enabled a 

 period to be carried back a considerable time before the 

 star was recognised as double, so that the two distinct 

 methods of observation have supplemented each other 

 to an important extent. 



Nor is this all. In several instances it has been 

 possible for Mr. Lewis to combine the discussion of the 

 meridian observations of the brighter component with 

 the relative motion of the fainter companion as deduced 

 from the micrometer measures, so as to gain an idea of 

 the relative masses of the two stars. The eighteen 

 cases with which he has been able to deal give some 

 very remarkable results. In twelve out of the eighteen 

 cases the fainter star has the greater mass; in three 

 cases the masses are about equal, although the stars 

 differ considerably in brightness; in three cases the 

 fainter star has the smaller mass, but the predominance 

 of the chief stju^ in brightness is enormously out of 

 proportion to its superiority in mass. " .'\s a rule the 

 apparent satellite is, in fact, the primary of the system." 

 The table which Mr. Lewis gives of these eighteen 

 stars is a most suggestive one, for it indicates that the 

 mass of the fainter star is much more closely related to 

 its colour, that is to say, to its spectrum, than it is to 

 its stellar magnitude. Given a star of the solar type of 

 spectrum with a companion purple or bluish in colour 

 (and, therefore, natur;dly assumed to be of the Sirian 

 type), and the latter is, generally speaking, of the 

 greater mass, though much inferior to the solar star in 

 brightness. 



Of course, the number of stars as yet available for 

 such treatment is far too small for any broad general 

 conclusions to be based upon them, but the importance 

 of Mr. Lewis's table from a theoretical point of view 

 can hardly be exaggerated. The natural way of look- 

 ing at stellar evolution is to assume that the star of 

 greater mass will cool the more slowly, and con- 

 sequently that, of the members of a pair, the heavier 

 .star must be the hotter; we also naturally assume that 

 it must be the more luminous as well. Vet the results 

 which Mr. Lewis has put before us, so far as ihey go, 

 emphatically negative this reasoning. Then, again, as 

 between stars of the solar and of the Sirian types, it 

 has been usually assumed that the latter are intrinsicallv 



much the brighter of the two, but in the view of the 

 facts here presented this assumption will in future have 

 to be revised. 



One of the early workers upon double stars, Sir 

 James South, on the publication of Struve's great work 

 in 1837, is reported to have exclaimed, "There is 

 nothing left for me to do ! " South had not learnt that 

 g-ood work does not preclude fiu'ther work, but forms 

 the best possible basis for it; and we see from Mr. 

 Lewis's Memoir at once how great and how important 

 has been the work which others have built upon the 

 foundation which W. Struve laid so broadly and so well. 



Liquid Air. 



The Knudsen System. 



.Since liquid air was made some years ago, at great 

 expense, in the laboratory, its possible applications in 

 industry as well as in science have become increasingly 

 evident, and at the same time new and improvea 

 methods of manufacture have steadily reduced its cost. 

 The latest and cheapest of these methods of manufac- 

 ture, which reduces the price of liquid air to a hun- 

 dredth, perhaps a thousandth, part of that at which it 

 stood a few years ago, is called the Knudsen system, an 

 installation of which has just been opened in Church 

 Road, Battersea, by Major B. Baden- Powell. Early 

 methods of liquid air manufacture depended on what is 

 known as the cascade system; that is to say, that con- 

 tinually decreasing temperatures were obtained by 

 causing successive volatile liquids to boil at reduced 

 pressures. Thus ethylene, which boils below the 

 freezing point of water, will boil at a lower temperature 

 when the pressure of the atmosphere is taken from its 

 surface, and will help to produce a degree of cold sulK- 

 cient to liquefy carbonic acid gas, while the normally 

 low temperature of liquid carbonic acid can similarly be 

 reduced by again reducing the atmospheric pressure. 

 Thus, as in a cascade, the temperature grows lower and 

 lower by successive fidls. The alternative method, of 

 which Knudsen 's is a type, depends on the principle of 

 the adiabatic expansion of gases. When a gas is com- 

 pressed heat is produced; when a compressed gas is 

 allowed to expand suddenly, cold is produced. 



In the Knudsen system atmospheric air is first of all 

 purified and freed from dust, and is then compressed in 

 a three-stage air compressor. In the first low pressure 

 cylinder it is compressed to 90 lbs. pressure, in the 

 second intermediate to 550, and in the third high 

 pressure cylinder the pressure is 2,500 to the square 

 inch. The air is cooled between the stages of com- 

 pression to the temperature of the cooling water. After 

 the air has been freed from dust and compressed to 

 2,500 lbs. to the square inch, it then enters a moisture 

 separator, which removes the moisture from the air, as 

 well as the oil and other impurities which might have 

 vaporised and passed over in the process of com- 

 pression. 



The air, still at a pressure of 2,500 lbs. to the square 

 inch, passes next through a number of copper coils. 

 Surrounding these coils cold air from tlie liquefier 

 passes in the opposite direction and lowers their tem- 

 perature sulliciently to freeze the traces of moisture on 

 the inside of the coils, leaving absolutely pure dry air 

 to be liquefied. 



