May 1, ia93.] 



KNOWLEDGE 



«)1 



and that streams of stars appear to spring from them and 

 to curve away into space. It is difficult to conceive of a 

 stream of stars that is not flowing. The alignment of a 



Fio. .3. — Untouched block made from the photograph of the Hercules 

 C'hister, taken at the Lick Obserratorv.which is reproduced in Fig. 2, 

 but black represents white, and the etching is not carried as far as 

 in Fig. 2. 



group of stars suggests their motion with diiferent velocities 

 from a common origin, and that they will be found to be 

 associated with a stream of matter. In the Pleiades cluster 

 we have actual photographic evidence of a narrow nebulous 

 stream of matter Unking together an alignment of six 

 stars ; and Prof. E.G. Pickering has photographed, in the 

 Orion region, a faint nebulous band of light which actually 

 links together a line or stream of sixteen faint stars,* and 

 many other instances of stars distributed along streams or 

 bands of nebulous matter might be quoted. 



Fig. 4. — Untouched block made from a photograph of the G-reat 

 Cluster in Hercules, taken by Dr. Isaac Roberts on 22nd May, 1887. 



In a space closely crowded with stars like the central 

 region of a star cluster we should expect frequent collisions 



to take place, and, as the result of such collisions, We 

 should expect to find streams of nebulous matter associated 

 with alignments of stars radiating from the central mass ; 

 in fact, we do find that nearly all star clusters show traces 

 of nebulosity, and that the brighter regions of most of the 

 great nebuliP coincide in position with clusters of stars 

 which the nebulous matter evidently surrounds. NebulsB 

 and star clusters are divided by no hard and fast line. 

 The larger nebuhe may be described as groups of stars 

 surrounded by bright nebulosity, and star clusters may be 

 defined as groups of stars surrounded by a faint nebulosity. 



I am indebted to Dr. Gill for the interesting photograph 

 of the ai Centauri cluster ; a general nebulosity will be 

 recognized extending through it, especially in the region 

 where the stars are most closely packed, and it will be seen 

 that the nebulous background is interrupted here and there 

 by narrow dark channels — similar to those we have recog- 

 nized in many nebulse, and in several nebulous regions of 

 the Milky Way. 



The blocks are made from several photographs of the 

 great cluster in Hercules — which have all been reduced 

 by means of photography to the same scale — and have been 



-4' -:5' 



+ 1' 



+ 3" 



+ 4' 



N 



* See "Annals of Harrard College ObserTatory," Vol. XXVII., p. 155. 



Fig. '). — Copy of a chart of the Great Cluster in Hercules, made by 

 Dr. J. Scheiner, of the Astro-Physical ObserTatory at Potsdam, in 

 1891. 



oriented with the same point uppermost, and turned over 

 ■where they were reversed right and left. The corre- 

 spondence between Dr. Isaac Roberts' photograph and those 

 taken by the Brothers Henry and at the Lick Observatory 

 will not at first be evident, because the exposure of Dr. 

 Roberts' photograph is considerably greater than the 

 .- exposure of the photographs taken at Paris and at the Lick 

 Observatory, and consequently the nebulous area shown 

 is larger, and all the stars in Dr. Roberts' photograph are 

 represented by larger patches than in the Paris and Lick 

 photographs. But making these allowances, it will be seen 

 that the photographs correspond, star for star, and that 

 most of the stars shown in Dr. Scheiner's chart can be 

 identified on the photographs. 



In the Paris and Lick photographs very curious pro- 

 jecting structures will be seen extending from the central 

 mass of stars and nebulosity. More than one of them 

 has the branching character which reminds one of the 

 forms taken by the streams of heated gas which rush 

 upward from the solar photosphere, but in the case of this 



