190 



KNOWLEDGE 



[October 1, 1891. 



Dr. Max Wolf, and there are similar dark tree-like struc- 

 tures to be traced on Prof. Barnard's photograph of the 

 Sagittarius region, a copy of which was published in the 

 •July number of Knowledge for 1890. Two of the most 

 remarkable of the dark branching structures in the 

 Sagittarius region spring from the dark area between the 

 two clusters of stars near the bottom of the plate, and 

 from this same dark area springs the bright tree-like 

 form referred to on page 51 (Knowledge, March, 1891) 

 as affording evidence of the projection of matter into a 

 resisting medium. 



It wiU be noticed that the dark channels and dark 

 branching structures referred to are all bordered by lines of 

 stars, which mark out their contour, just as the dark areas 

 around -t, Argus are bordered by lines of stars. There are 

 three striking dark patches, irregularly bordered by stars, 

 near to the edge of the upper portion of the bright cloud- 

 like mass to the right hand of the 'j. Cygni pictm-e. They 

 are best shown in the large plate, though they are just 

 visible in the small one. These dark patches remind one 

 of the small dark hole bordered by stars in the Sagittarius 

 region (see Knowxedge, July, 1890, p. 17o). The lowest 

 of the dark patches on the Cygni plate has an elliptic- 

 shaped group of small stars near its centre, surrounded by 

 a dark elliptic channel. The patches seem to be connected 

 together, and to form part of a dark structure, springing 

 from the dark area to the south of y. Cygni. 



Fia. 4. Lightly etched plate, showing ciu-rmg ntbul;i 

 and nebulous branches from 52 Cvgni. 



A remarkable series of slightly curved lines or strings 

 of stars will be seen near to the top of the e Cygni 

 plate. The plate should be held sideways to see them 

 best, as they run nearly vertically, that is, north and south 

 upon the plate. I have made an enlarged photograph of 

 this region, which I have had etched upon a zinc block 

 (Fig. 3), but as many of the small stars are lost, the 

 lines of stars are not as strikingly shown as upon the plate. 

 It, however, shows a very curious httle circle of stars, 

 ■\vith radial streams, near to T Cygni. The radial hnes of 

 stars are evidently connected with the Uttle circle of stars 

 from which they radiate. The word "circle " is printed just 

 above the small circle of stars on the block, but its striking 

 character is best recognized on the plate. 



That these curves and allineations of stars are not 

 fanciful forms or chance arrangements, such as might be 

 detected in the grouping of small objects thrown down at 

 random (such as drops of rain falling on a paving stone), 

 will be evident directly we consider the question from a 

 probability point of view. Many of these star-streams 

 contain twenty or thirty or more stars of about the same 

 magnitude, following one another in a straight or curving 

 line, at approximately equal intervals. If we assume that 



the chance is one to four that a point thrown down at 

 random would appear to fall into line with two points 

 already in position, the chance against ten such points 

 falling into line with two already in position, so as to form 

 a smooth curve, would be more than a million to one, and 

 the chance against twenty such points falling into line would 

 be more than a million of millions to one. This is much 

 understating the improbability, for leaving out of consi- 

 deration the chance that a series of adjacent stars having 

 no physical connection should all be of nearly the same 

 magnitude, and at about equal spaces from one another, 

 the chance that the next adjacent star should fall into line 

 with a series already in a straight or curving row must 

 be less than yVth, for the eye would certainly detect 

 irregularities or deviations from the general trend of the 

 curve amounting to 30°. We might, therefore, without any 

 further evidence, take it as established that there must be an 

 intimate connection between the stars of such a stream, and 

 that they must have had a common origin, and form a system. 

 But in the Pleiades cluster we have ocular evidence of 

 a physical connection between the stars forming two such 

 streams. The individual stars are connected together by 

 a narrow nebulous band, " which threads them together." 

 Prof. Pickering has also succeeded in photographing a 

 faint nebulous band of light in the Orion region which 

 passes through sixteen faint stars.-' We have also some 

 endence of a similar connection on Dr. Max Wolfs 



-!<.*■. 



Fig. o. Deeply 

 cuiTing nebula. 



etched block, showins stars in\olved 



; Cygni photograph : I have had the small etchings, 

 Figs. 4 and 5. made from a part of this plate, which shows 

 a curving nebula involving stars. Fig. 4 has been lightly 

 etched so as to show the nebulous band, and Fig. 5 has 

 been more deeply etched so as to show the stars involved 

 in the nebula and linked together by it. Fig. 4 also 

 shows traces of two nebulous branches from the star 

 52 Cygni. It will be .seen that small stars lie along these 

 curving branches and appear to be connected together by 

 them. 



ILcttcrs. 



[The Editor does not hold himself responsible for the opinions or 

 statements of correspondents.] 



THE OBSERVATION OF BED STAE8. 

 To the Editor of Knowledge. 



De.vk Sir, — I fully agree with ^Ir. Franks that no 

 demonstrative proof of colour instability in non-periodical 

 stars is at present forthcoming. Nor is his scepticism 



• See "Annals of Harvard College Obserratory," vol. xivii.. p. 155. 



