Total Eclipses of the Sun. 417 
to the subject, and to discuss more fully the details contained 
in recent reports upon this matter, without, however, touch- 
ing upon the fundamental principles set forth at page 8 of my 
Yearly Report for 1854, with which I assume the reader to be 
familiar. 
At the outset I must remark that when in a dark room an 
opening in a window-shutter, a! b' (Plate I. fig. 1), is observed 
from ¢c through a film of vapour A B, the visibility of the vapour 
depends on the brilliancy of the entering light*. If the bril- 
liancy is very considerable, the light from a6 falls on the whole 
breadth of de, and is reflected to the eye from ail points of the 
vapour, thus rendering the entire film of vapour visible, though 
with very unequal inteusity. The intensity will diminish from 
f to d and from g to e, because the angle of incidence and re- 
flexion is more and more inclined ; the greatest intensity will be 
manifested in the space fg, where the light passes from the 
opening direct through to the eye. The more the brilliancy of 
the light that enters at the opening in the shutter diminishes, 
the more must the visibility of the vapour decrease, beginning 
from d and e; anda limit is soon reached where the side portions 
fg and de disappear altogether, the part fg alone remaining 
visible. We see from this, that when a surface of light of very 
low intensity is observed through a mass of vapour, only that 
portion of the vapour can be perceived which is between the eye and 
the luminous surface. 
If the opening ad is so disposed that the light beginning at a 
goes on decreasing towards 8, then in like manner the visibility 
of the vapour will diminish from f towards g, and will only be 
extended over a certain portion of the space fy. 
To render what has been stated applicable to the phenomenon 
of total solar eclipses, it is requisite, in the first place, to form 
a precise notion with regard to vapour. On examining carefully 
the formation of clouds, especially as they frequently present 
themselves to us in summer, it will be remarked that the vapours 
which congregate together form no regular covering, but a ragged 
interwoven mass (fig. 2), consisting of thicker parts, and parts 
growing less dense till they vanish altogether. It will further 
be noticed that very frequently the vapour disappears at one 
side or at a corner—melts away as it were—and is deposited on 
another side. 
As far, moreover, as the light is concerned which serves for 
rendering objects visible during total solar eclipses, it consists of 
* Ifan uncoated looking-glass is besprinkled with alum-water and then 
allowed to dry, a fine coating is formed thereon, on looking through which 
the same appearances are presented as through a mass of vapour. This 
arrangement is suitable to the experiment in question. 
