8 Drs Christison and Turner on the Construction of 
ingenious apparatus of Count Itumfordj and we know that some, whose re- 
sults have been a good deal relied on, employed the crude and inaccurate me- 
thod of estimating the intensity of the shadows, as they were simply cast on a 
white wall in an open room. It is not at all surprising, that experiments con- 
ducted in a manner so unscientific have disagreed with each other, and brought 
discredit on the whole method of investigation. 
The remarks we have now to make, on the accuracy of the indications pro- 
cured by the comparison of shadows, must of course be understood to apply 
only to experiments made with a due regard to every source of fallacy ; in 
short, to the photometer of Count Iiumford. We need not describe this in- 
strument. It will be sufficient to mention, that the chief advantages of its 
construction are, that it secures the uniform equality of the angles of the inci- 
dent rays, and protects the eye from every light except what illuminates the 
shadows, and a very small space around them. 
The objections that have been publicly made to this mode of experiment- 
ing are three in number. 
First, it is said, that the eye cannot judge with adequate precision of the 
relative depth of the shadows. This objection is a valid one to the rough 
method of experimenting adverted to above, especially when the observer 
happens also to have an inaccurate eye. But it is quite inapplicable to 
Count Rumford’s apparatus, in the hands of a person with a tolerably cor- 
rect eye. We have found, that even those altogether unaccustomed to 
scientific experiments could easily distinguish, after a few trials, a diffe- 
rence of a fiftieth part between two lights ; and that, when one of us had 
adjusted the lights to his satisfaction, every other person in the room (to the 
amount sometimes of four or five), uniformly agreed with him as to the identity 
of the shadows. On several occasions, too, more particular facts have occurred, 
which prove, beyond a doubt, the extreme delicacy and correctness of the eye 
in such experiments. Thus, on comparing two gas-jets of the same size with each 
other, taking care to avoid all means of prejudging the distances, the result of 
the calculation gave a ratio of 100 to 101.4. And, again, when the light of a 
wax candle was twice compared with a gas-jet, of the same size, several days 
having intervened betwixt the observations, and the distances being purpose- 
ly altered, the ratios were 58 and 58-7 to 100. Similar occurrences were com- 
mon. These statements accord very nearly with the experience of Mr 
Nicholson, who found he could detect a difference of an 80th part between 
two lights. 
But it has also been objected, and with a greater shew of reason, that, al- 
though the method is accurate when applied to lights of the same nature, it is 
not so when they differ much in colour ; because the colours of the shadows differ, 
and the eye cannot abstract the difference of colour from the difference of 
shade. 
Our experience on this point is, in our opinion, quite decisive. If the dif- 
ference of colour is slight, it constitutes no impediment. If greater, it will 
lead to a small error, unless the observer take care to try the effect of shift- 
ing the moveable light an inch or more on each side of the point at which he 
supposes the shadows equal. But in this way he will at length hit on the ex- 
act distance. If the difference, however, is very great, little or no advantage 
