60 Douljle Stars — Colours of Stars. 



for his micrometer-lamp, with so much advantage that he 

 could never tolerate green again. And in the same manner 

 Dembowski finds that a vermilion illumination has " un effet 

 surprenant," — stars even of the three smallest magnitudes of 

 Struve's scale, barely visible in a dark field, and entirely- 

 effaced by the approach of the illuminated threads, becoming 

 not only visible, but in general even measurable, upon a red 

 background. That the effect of contrast, in bringing out the 

 blue light so frequent (as we shall hereafter remark) among 

 minute comites, is not concerned in this curious result, is evident 

 from the Baron's further remark, that blue is next in good 

 effect to red, yellow least favourable of all. 



The Baron has noticed a source of discordance in his 

 measures of position, which it may be useful to mention. He 

 has found the values differ very sensibly according to the 

 direction in which the position-circle is made to revolve, and 

 hence takes a mean of the results obtained in each direction. 

 The deception is only of occasional recurrence, and may be an 

 individual peculiarity ; the supposition, however, that other 

 eyes may be similarly affected, would account, he observes, for 

 some discrepancies in the measurements of position by dif- 

 ferent observers at the same epoch. 



The perfection of his optical means may be estimated from 

 his having repeatedly 'measured the distance of the close pair, 

 or technically speaking A and B, of £ Gancri, during 1863, 

 giving at a mean about f ". Those among our readers who 

 are acquainted with this object will readily comprehend the 

 difficulty of the feat. In general, however, his values under 

 1" are only estimations. 



The Baron's observations, which will no doubt be pub- 

 lished, as heretofore, from time to time, will be looked for with 

 the greater interest, as the results of the measurements of a 

 very large number of double stars taken from Struve's class of 

 " lucidee," during the past year, with the Oxford heliotneter, 

 have not borne out that astronomer's impression as to their 

 binary character. The modes of measurement, however, are 

 essentially different — the heliometer values being obtained 

 from the doubled image produced by a divided object-glass, 

 the halves of which are relatively moveable, those of Dem- 

 bowski and Struve from the parallel-thread micrometer. 



COLOUKS OF STAES. 



We are induced to offer a few additional remarks upon this 

 curious subject, that it may be presented in a less incomplete 

 form before our readers. 



As the susceptibility of different eyes varies both with re- 

 gard to the intensity and colour of light, it must be expected 



