SUCCESSIVE IMPROVEMENTS. © 267 
which are horizontal, directed towards the land, have 
only been produced to be entirely wasted. 
This horizontal beam of rays not only forms a very 
small part of the total light ; it has also the serious incon- 
venience of diminishing in intensity as it diverges, and of 
not extending itself to a distance without being sensibly 
enfeebled. To destroy this unfortunate loss of light,—to 
profit by all the light which the lamp emits,—was the 
twofold problem which remained to be resolved in order 
to extend the range, and thus the utility, of light-houses. 
Concave metallic mirrors, called parabolic reflectors, 
have furnished a satisfactory solution. 
When the lamp is placed at the focus of such a mirror, 
all the rays which emanate from it are brought, by the 
reflexion they undergo against its sides, into a common 
direction ; their original divergence is destroyed ; they 
form, as they issue from the apparatus, a cylinder of light 
parallel to the axis of the mirror. This beam is trans- 
mitted to the greatest distances with the same brightness, 
except that the atmosphere absorbs a small part of it. 
Before proceeding further, let us stop to observe that 
this construction is not without an inconvenience. We 
thus indeed easily bring to bear on the horizon of the sea 
a multitude of rays which would otherwise have been lost 
on the ground, in space above, or on the side towards the 
land ; and we overcome the divergence of those rays 
which would naturally be directed towards the navigator. 
But the cylinder of reflected rays can have no greater 
breadth than that of the mirror; the space which it illu- 
minates has precisely the same breadth at all distances ; 
unless indeed we employed many similar mirrors, pointed 
different ways, and even then the horizon would include 
many large spaces completely dark, in which the pilot 
