ch 
222 Prof. Draper's D tion of the Tithonometer. 
From this it will be perceived that, taking the first experiment | 
as an example, if at the end of 3t i P 
700, at the end of 60” it has r 0 
90”, 7-50 more, at the end of 120”, 7: 
each thirty seconds. And, when it is recollected that the read-_ 
ings are all made with the instrument im motion, the differences _ 
between the numbers do not greatly exceed the possible errors of 
observation. It may be remarked that the hird and fourth ex- 
periments were made with a different lamp. 
idescent as that here used will pass the lens, its ef 
never be mistaken for those of the tithonic rays. This ~ 
is easily understood, when we remember that the effect of such ' 
_ transmitted heat would be to expand the gaseous mixture, but — 
the tithonie effect is to contract it. 
Next, I shall proceed to show that the indications of the titho- 4 
nometer are strictly proportional to the quantity of rays that have — 
_ impinged upon it; a double quantity producing a double effect, a 
triple quantity a threefold effect, &c. q 
A slight modification in the arrangement (fig. 4) enables us to 
prove this in a satisfactory way. The lens D, being mounted in 
a square wooden frame, can easily be converted into an instru- 
ment for delivering at its focal point, where the sentient tube is 
placed, measured quantities of the tithonic rays, and thus becomes 
an invaluable auxiliary in those researches which require known 
and predetermined quantities of tithonicity to be measured out. 
The principle of the modification is easily apprehended. If half 
the surface of the lens be screened by an opake body, as a piece 
of blackened card-board, of course only half the quantity of rays 
will pass which would have passed had the screen not been in- 
terposed. If one fourth of the lens be left uncovered, only one 
fourth of the quantity will pass; but in all these instances the 
focal image remains the same as before. By adjusting, therefore, — 
upon the wooden frame of the lens, two screens, the edges 0 
which pass through its centre, and are capable of rotation upon 
that centre, we shall cut off all light when the screens are applied 
edge to edge, we shall have 90° when they are rotated so as to 
be at right angles, and 180° when they are superposed with their 
edges parallel. Thus by setting them in different angular posi- 
4 
