210 



NA TURE 



[July 2, 1908 



provide and remove fear of defects in this direction. 

 When to this is added the risi< introduced by every screw 

 passing through the copper, such as, for example, those 

 necessary for the fixation of the bed, the balance beams 

 from which the radiator system is suspended, &c., there is 

 at present fear enough. It was in indicating points such 

 as this that Prof. Benedict's visit was of so much value. 



The heat produced within the copper box, carefully pre- 

 vented from escaping from its surface in the manner 

 described, is carried out in a stream of water constantly 

 driven through the radiator system. The quantity of water 

 passing is varied with the necessities of the moment, and 

 is necessarily much greater when the subject is at work 

 on the ergometer. The water passing is weighed on exit 

 in a meter, and its temperature on entrance and exit 

 observed. The former is practically constant, the latter 

 kept as constant as possible by variation in the rate of 

 water flow. These data form the main items in the state- 

 ment of heat produced, though other important items, such 

 as the amount of water condensing on the radiator pipes 

 within the chamber, the water evaporated and carried 

 through with (he air leaving the chamber, are dulv con- 

 sidered and provided for. Thermocouples are placed in the 

 tubes through which the air finds entrance and exit, in 

 quantity 7-; litres per minute, so as to ensure the detection 

 of any difference of temperature. Any difference found 

 is corrected by increased heating of the entering air. 



In Benedict's calorimeter the air is driven from the 

 chamber through a closed system of tubes back to the 

 chamber again. In this system are interpolated sulphuric 

 acid and soda lime absorbers of a necessarily very large 

 size, determined by the large mass of air in motion per 

 unitof time. The oxygen consumed is made good by 

 admission of oxygen from a cylinder of the compressed 

 gas. In the Sheffield apparatus this will not, at first at 

 least, be attempted. Dried air heated and passed into the 

 chamber will be driven out through a sulphuric acid 

 absorber, no attempt being made to do more than take a 

 determination of the water. 



RECENT RESEARCHES IN THE STRUCTURE 

 OF THE UNIVERSE.' 



T CONSIDER it an uncommon privilege to lecture on 

 the structure of the universe in the country of the 

 Herschels. Even now their celebrated gauges are un- 

 rivalled, and they still form one of the important elements 

 on which any theory of the stellar system must be based. 

 It is well known that the plan of these gauges consisted 

 in directing the telescope successively to different points 

 all over the sky, and simply counting the number of stars 

 visible in the field. 



There is one fact clearly brought out bv these gauges 

 to which I must direct your attention. It'is that in the 

 outward appearance of our nightlv skv, as seen with the 

 telescope, there is a great regularity, 'in the Milkv Way, 

 that belt which wo see with the naked eye encircling the 

 rvhole of the firmament nearly along a great circle, the 

 number of stars, as seen in Herschel's 20-feet reflector, is 

 enormous. On both sides this apparent crowding of the 

 stars diminishes very gradually and regularly until, near 

 the poles of the Milky Wav, we come to the 'poorest parts 

 of the sky. 



Let us look at this phenomenon somewhat more closely. 

 If we direct our telescope first towards the part of the 

 Milky Way near Sirius, and if from there we gradually 

 work up towards the North Pole of the Milkv Way in 

 the constellation called the Hair of Berenice," we shall 

 clearly perceive this gradual and regular change in the 

 number of stars. Now if we repeat the same process, 

 beginning from some other point of the Milkv Wav, say 

 in Cassiopeia or the Southern Cross, we shall find that, 

 not only is there a similar gradual change, but we shall 

 approximately go through the same changes. 



At the same distance from the Milkv Wav we shall 

 find, approximately, the same number of stars in the field 

 of the telescope. Put in other words, the richness of stars 



1 Discourse delivered at the Roy.il Institution on Friday, May 2j, by 

 Frof. J. C. Kapteyn. 



NO. 2018, VOL. 7SI 



varies regularly with the galactic latitude ; ' it varies 

 relatively little with the galactic longitude. 



Imitating most of the investigators of the stellar system, 

 we will therefore disregard the longitude and keep in view 

 only the changes with the galactic latitude. In reality 

 this comes to being satisfied with a first approximation. 

 For, in reality, there are differences in the different longi- 

 tudes, especially in the Milky Way itself. But even here 

 the differences are not so great as seems commonlv to 

 be supposed. There is every reason to believe, therefore, 

 that our approximation will be already a tolerably close 

 one. 



Real Structure. 



Meanwhile, what the Herschel gauges teach us is only 

 relative to the outward appearance of the skv. What is 

 the real structure of the stellar world? If we see so 

 many stars in the field, with the telescope directed to the 

 Milky Way, is it because they are really more closely 

 crowded there, as Struve thinks, or is the view of the 

 older Herschel correct, who imagined that the greater 

 richness is simply a consequence of the fact that we are 

 looking in deeper layers of stars ; that our universe is more 

 extensive in the Milky Way than it is in other directions? 



Imagine that we could actually travel through space. 

 For instance, imagine that first we travel in the direction 

 of the constellation Cassiopeia. If we travel with the 

 velocity of light, not so very many years would pass before 

 we get near to some star. Proceeding on our journey for 

 many, many more years, always straight on, we will pass 

 more stars by and by. How will these stars look thus 

 viewed from a moderate distance — say, from a distance as 

 that of the sun ? Will they all be found to be of equal 

 luminosity, as Struve practically assumed? And in this 

 case are they as luminous as our sun, or more so, or less 

 so? Or are they unequal? If so, how many of them 

 are brighter than our sun, how many fainter? Or, to 

 be more particular, how many per cent, of the stars are 

 10, 100, 1000, &c., times more luminous than our sun? 

 How many are equal to the sun, or 10, 100 times fainter? 

 In a few w-ords : What is the nature of the mixture? or. 

 lastly, what is the mixture laxv of the system of the stars? 



Furthermore, in travelling on, shall we find the stars in 

 reality equally thickly, or rather thinly, crowded every- 

 where.' Or shall we find that after a certain time, which 

 iTiay be many centuries, they begin to thin out, as a first 

 warning of an approaching limit of the system? Is there 

 really such a limit, which, once passed, leads us into 

 abysses of void space ? 



Herschel thought there was such a limit, and even 

 imagined that his big telescope penetrated to that limit ; 

 that is, he assumed that his telescope made even the 

 remotest stars visible. On this supposition is based his 

 celebrated disc theory of the system. 



Again, we may condense these questions in this single 

 query : How does the crowding of the stars, or the star- 

 density, that is, the number of stars in any determined 

 volume (let us say in a cubic light century), vary with the 

 distance from our solar system ? 



But there is more. We supposed that our journey went 

 straight on in the direction of Cassiopeia, which is in the 

 Milky Way. What if our journey is dir;ected to the 

 Pleiades, which are at some distance from that belt, or 

 to the Northern Crown, which is still further, or to the 

 Hair of Berenice, which is furthest of all from the Milkv 

 Way.' For different regions equally distant from the 

 galaxy we have seen that outward appearances are the 

 same. We may admit, with much probaBility, that in 

 space, too, we would find little difference. Summing up, 

 the problem of the structure of the stellar system in a 

 first approximation comes to this : — 



To determine, separately for regions of different galactic 

 latitude, in which way the star-density and the mixture 

 vary with the distance from the solar system. 



I think that there is well-founded hope that, even 

 perhaps within a few years, sufficient materials will be 

 forthcoming which will allow us to attack the problem to 

 this degree of generality, with a fair chance of success. 

 .At the present moment, however, our data are yet too 

 scanty for the purpose. Still, they will be sufficient for 

 the derivation of what must be in some sort average con- 



