72 CALIFORNIA ACADEMY OF SCIENCES. [Proc. 3D Ser. 



lines could be seen moving swiftly lengthwise across the 

 illuminated lens in the direction of the wind. 1 



At San Francisco, in the strong "wind gap' of the 

 Golden Gate, it sometimes happens at sunset, after a calm, 

 moderate day, that star images are disc-like and steady; in 

 a few minutes the westerly wind rises and brings in the fog- 

 chilled air from the ocean, with flying patches of fog over- 

 head, and the star images at once become diffused and 

 nebulous, with an unsteadiness reaching an amplitude of 

 5" to 10" of arc. 



Large volumes of air, each of different but uniform tem- 

 perature, projected across a line of sight at a great distance 

 from the observer, do not necessarily produce abnormal 

 images of a star or heliotrope; they seem to act as prisms, 

 and deflect the line of light so as to exhibit abnormal hori- 

 zontal or vertical refraction without any marked unsteadi- 

 ness. Along the face of a great mountain wall we have 

 measured a slow change of azimuth of a signal amounting 

 to 65" of arc, due to the gradual rising of the morning tem- 

 perature of the air immediately adjacent to the eastern rocky 

 face of Mt. Constitution. 



There are so many conditions conspiring to a disturbed 

 atmosphere that it necessarily acts in various ways difficult 

 to predict. Where there are regular, -irregular or confused 

 atmospheric waves moving at different distances from the 

 objective, and in different directions, they necessarily give 

 different images of the star at different foci, and at each 

 change they bodily shift the image near the focus, and also 

 change its form. Professor Douglass' observations are very 

 instructive in this matter. 



When volumes of unusually heated air pass over the 

 objective they may act as a species of air lens which sud- 

 denly and irregularly spreads out the image of the star at 

 the focus; but no change of focus will bring the image to 

 condensation. This we first experienced at Point Concep- 

 cion in 1850, where the perceptibly hot volumes of air from 



1 Douglass, " The American Meteorological Journal," March, 1895, p. 395. 



