Chemisiry and Physics. 409 
mirror with directive mounting—has been one of our regular field 
instruments in triangulation for seks years, and has been used on 
lines from 20 to 192 miles 1 in lengt 
the annual report of ‘ASaotaind George rive redtie in charge of the 
The stations named are along the western crest of the Sierra 
Nevada Mountains, along the crest of the immediate Coast Range, 
and Mt. Shasta, at ‘the head of the Sacramento Valle 
cess Was per 
rele de |—Heliotrope Spectra. 
My former experience of the decomposition of the rie cs 
image of the sun after passing through many miles of the atmos- 
phere was fully verified. The heliotrope images were midon 
decomposed on the lines under seventy miles in length, but they 
were, as a rule, decomposed on all — vee and ranged 
Blue Gre Gre 
through the formule sheng Yellow Yellow Orange the last 
R Red ‘Red. 
being the less frequently seen. The peculiar sparkling character- 
istic of the b bright heliotrope image does not announce itself in 
the spectrum image; the colors give a steady, soft t image, which 
is generally slightly higher than broad; frequently twice as high 
as broad; and upon some “gig it reaches a height of 60 sec- 
‘nae with a breadth of 10” to 15’. 
Tn nearly all cases the se is the most marked and is certainly 
the most persistent ; the blue will fade away, the yellow or orange 
is not in sufficient contrast with the white field to be observed 
bie The color of the red is that of the spectrum at the B line. 
‘requently the red will exist with a sharply defined nucleus of 
t orange, or even white; but it is doubtful if this exists when 
the spectrum i d. 
e column of spectrum light sometimes undergoes the most 
unceasing apparent interchange of the colors, as if the rapid 
chan nges in vertical refraction suddenly shortened and lengthened 
the column; and yet tests failed to show that any part of the col- 
umn was not in oti vertical. 
