IS 



KNOW LI.DGi:. 



Jamarv, 1912. 



i|uili; ciilaiii tli.ii .ill tlu l>ri(,'litcT nu-mlx-rs liavu 

 now Ih-i'M (liscovcnHl, tliouf^li fainter onrs from 

 titiu- to tinu- an- (li-ti-cti-d. Now. if tlio stars of 

 type \' are plotted ilown, it is found that they 

 coincide in a riMuarkahle manner with the central 

 circle of the Milky Way; s<» nuidi so, indeed, 

 is this the case that tlu' circle is clearlv marked 



FlGCRE 34. Sagitla. 



nut In' them when they are sufficiently numerous, 

 and even where the f^reat hifurcation takes 

 ])lace, these objects persistentlx" follow the centre. 

 The average deviation of the whole 

 number is probably less than twn 

 degrees. .As .Miss Clarke has ver\' truK 

 said, we must belie\e that the Milk\ 

 W'av is very much as we see it, and 

 that it is not to be explained by two 

 intersecting rings or by spirals. The 

 \'-t\"pe stars may then be deemed to 

 mark the absolute equator of the Uni- 

 verse of stars. Now in the Southern 

 Hemisphere, near the poles, and (]uite 

 separated from all apparent connection 

 with the Milkv \\'ay. are two cloud-like 

 masses called the Nubecula Major and 

 Minor. Herschel. after examining the 

 Nubecula Major, was clearly of opinion 

 that it was a iiiiilttiiii in parvo of the 

 visible universe. Here, in the space of 

 a few degrees, are found stars, irregular 

 and globular clusters, nei)ulae planetary 

 and irregular, all commingled in inex- 

 tricable confusion. Photographs of the 

 spectra of the objects therein contained, 

 made with the twenty-four-inch Bruce telescope, 

 have confirmed Herschel's dictum by re\ealing 

 twentv stars of Type W fifteen objects showing the 

 spectra of gaseous nebulae, and eight Orion stars 

 with bright lines. — Comparisons of the |)lates have 

 further revealed eight hun<lred and eight variable 



stars ! — Hut these new facts may give us a clue b\ 

 wiiich the seeming inextricable confusion may Ix; 

 reduced to some sort of appan-nt order. If the 

 objects showing bright lines are charted, several 

 facts are immediatel\ obvious. There are first 

 of all distinct grou|)ings, the most remarkable 

 being that connected with the great nebula. 

 30 Doradus. In this group, in a circle 

 of one degree, there are six stars of 

 Ty])e \', three gaseous nebulae in- 

 I hiding .50 Doraclus, and five bright 

 line Orion stars. In the same wav 

 the position of N.G.C. 1763 is marked 

 by three stars of Type \'. and four 

 gaseous nebulae. Now. if the Nubecula 

 Major really is a miniature universe, 

 there is a strong ])robability that the 

 stars of Type \' mark its ecpiator. 

 Obviously if the Nubecula was situated 

 at right angles to the line of sight, the 

 stars of Type V would tend to arrange 

 themselves in a circle : if, on the other 

 hand, it was inclined to the line of 

 sight, the form would be an ellipse, the 

 eccentricit\' varving with the inclination. 

 To test this, a chart was made of the 

 Niii)eciila and affixed to a wall, and 

 illiiniinated by a light jjlaced some 

 tweUe feet away : then between the 

 two was placed a circle of cardboard 

 mounted on a tall stand, and movable round an 

 axis. Repeated trials showed that it was 

 possible to draw an ellipse, with the following 



FlGUKli 35. I'erseus. 



results for stars of Tyjie \' : — 



Tot.il number ... ... ... ... -0 



Nc.ir ellipse ... ... ... ... 14 



Inside ... ... ... ... ... i 



Outside 3 



The ellipse was then roughly marked on the 



