PORTION OP THE COSMOS. SCINTILLATION OF STARS. 67 



8 -feet heliometer. The determination of the limits of teles- 

 copic visibility of small stars during daylight, in different 

 climates and at different elevations above the level of the sea, 

 has both an optical and a meteorological interest. 



Among the phenomena belonging to natural and tele- 

 scopic vision which are remarkable in themselves, and of 

 which the causes are much contested, is the nocturnal 

 sparkling (twinkling or scintillation) of the stars. According 

 to Arago,( 129 ) there are two things to be essentially distin- 

 guished in the scintillation : 1. alteration in the intensity of 

 the light by a sudden decrease, amounting even to extinc- 

 tion, and rekindling ; 2. alteration of colour. Both altera- 

 tions are even stronger in reality than they appear to the 

 naked eye; for when the several points of the retina are 

 once excited, they retain the impression of light which they 

 have received; so that the disappearance of the star, its 

 obscuration, and its change of colour, are not felt by us in 

 their full measure. Still more striking is the phenomenon 

 of scintillation seen through a telescope, if the latter is 

 shaken. Fresh and fresh points of the retina are excited, 

 and there appear coloured and often interrupted circles. In 

 an atmosphere composed of constantly varying strata of dif- 

 ferent temperature, moisture, and density, the principle of 

 interference explains how, after a momentary coloured flash, 

 there follows an equally momentary disappearance or obscu- 

 ration. The undulation theory teaches in general that two 

 rays of light (two systems of waves) proceeding from one 

 source (one centre of vibration) by the inequality of their 

 paths destroy each other; that the light of one ray, added 

 to that of the other ray, produces darkness. When one sys- 



