438 SCIENTIFIC RECORD FOR 188i. 



given iu the tables; the surfaces of equal density produce a certain 

 variation from the normal refraction given in the tables; the surfaces 

 of equal density being, as a rule, inclined to some degree instead of being 

 horizontal, and the degree of inclination being submitted to a certain 

 periodicity during a whole year, there necessarily arises from this cause 

 a certain correction to be applied to the position of a star, much like that 

 of the annual parallax and aberration, and which might be described as 

 "parallax of refraction." As this correction must obviously afiect the 

 values of the annual parallax and of aberration, it is easy to understand 

 the necessity of determining its true value with much accuracy. The 

 values deduced by Glasenapp for the stars of i UrssB Majoris, i and 6 

 Draconis, are — 0."04, — 0."11, and — 0."11, which figures would explain 

 to a certain extent tlfe negative parallaxes received by M. Nysen ("Nu- 

 tation derErdaxe"), and which, respectively, are — 0."03 — 0."05, and 

 — 0."06. The whole work of M. Glasenapp on this subject will soon be 

 published. {Nature, xxiii, p. 373.) 



Professor Tait communicates to the Eoyal Society of Edinburgh a 

 valuable communication on the subject of mirage. He has discovered the 

 simple assumption as to the law of the variation of density in a stratum 

 of air near the earth's surface which is necessary and sufficient to ex- 

 plain the simultaneous appearance of erect, inverted, and again erect im- 

 ages which have been frequently observed, but never before explained. 

 {Nature, IXV, p. 92.) 



Tait, at the subsequent meeting of the Eoyal Society of Edinburgh, 

 June 5, communicated the second part of his paper on mirage, in which 

 he proceeded to investigate more carefully that distribution of atmos- 

 pheric pressure which would explain the mirage phenomena. 



Two horizontal strata of uniform but different densities, separated by 

 a stratum whose density varies continuously from the one to the other, 

 were found to give results in close agreement with observation. That 

 a stratum of air should remain of i^ractically uniform density, through 

 even a comparatively small height, requires a lowering of temperature 

 to compensate for the diminution of pressure as the height increases; 

 but this rate of change of temperature, Professor Tait showed, was not 

 greater than had been observed in balloon ascents. 



With given thicknesses of strata there was a critical minimum dis- 

 tance at which mirage could be obtained. For greater distances there 

 were three images, two direct and one inverted. 



The inverted one was always larger than the lower direct one, but 

 only appreciably so when the distance of the object approached this 

 critical minimum value, for which the phenomenon known as "looming" 

 became e\ident. The second direct image is usually much the small- 

 est, being, except at distances near this same critical distance, so small 

 as to be practically invisible. This seems fully to account for the com- 

 paratively few instances iu which the three images have been observed. 

 Multiple inverted images, as observed by Scoresby, were explained as 



