blue. The period of its duration too is likewise variable. 
Sometimes it lasts but half a second, ordinarily perhaps a 
second and a quarter, and occasionally as much as two 
seconds and a half 
When examined with the assistance of a telescope it be- 
comes evident that the green ray results at a certain stage 
of the solar obscuration, for it begins at the points or cusps 
of the visible segment of the sun, and when the setting” is 
nearly complete extends from both cusps to the central 
space between, where it produces the momentary and in- 
tense spark of coloured light visible to the unaided eye. 
From the fact of the green cusps being rounded I appre- 
hend that irradiation contributes to the apparent magnitude 
of what is seen. The range of colour too as seen in the 
telescope is more varied, and the duration of the whole phe- 
nomenon more extended, than when the observation is made 
only with the naked eye. 
Of the objective nature of the phenomenon it is needless 
to offer evidence ; for it needs to be but seldom seen to pre- 
clude the idea of an optical illusion. That the waters of 
the ocean have nothing to do with the production of the 
colour is made manifest by its visibility when the sun sets” 
behind the edge of a well defined cloud. On the 14th and 
15th of June, for instance, it was seen at upper contact of 
the solar limb with clouds. On the earlier date in question 
a thin band of cloud stretched across the setting sun, and 
under a power of fifteen diameters the green effect was seen 
at upper contact with the cloud and again at final disappear- 
ance below the horizon. On the later date it was again seen 
at upper contact with each of several filaments of cloud 
and again at final disappearance. And on several other 
occasions the writer has observed the effect when the dis- 
appearance of the sun has taken place at an elevation of six 
or eight degrees behind a heavy bank of clouds. 
Respecting the increased range of colours seen when the 
