58 W. A. Rogers—New Diffraction Ruling Engine. 
(d.) In the investigations of wave lengths from a given ruled 
grating there are certain requirements which seem to need ex- 
perimental as well as theoretical researc 
First. It is important to ascertain the ‘best relation between 
the width of the lines and the width of the intervening spaces. 
In this machine, after having obtained a good line of a given 
width, the width of the spaces can be varied at wi 
Second. All we can say of the wave-length equation A=e sind, 
is that it expresses the length of a single wave of light of a 
given color. When we take into consideration the conditions 
under which the waves reach the eye, it is at least an open 
question whether the form of the uation is not somewhat 
more complex. As an illustration, the law of refraction in 
passing from one thin stratum of atmosphere to its adjacent one 
is quite simple, but it does not follow that this law holds good 
for the combined strata which make up the atmosphere “at a 
given altitude above the horizon. 
In order to ascertain the effect of errors of any class upon 
the position of the solar-lines it seems important to determine 
the effect of these errors experimentally. Hence the machine 
as been constructed in such a way that it is possible’to intro- 
duce at any point, either any single error of a given magnitude, 
or any combination of errors, without interfering with the 
remaining graduations. — 
e tremors communicated to the machine by the running- 
rhachinery of the fantory prevent the best work; but the ex- 
periments already made justify the hope that when it is per- 
manently mounted upon the firm foundation already prepared 
for it, work may be done which will contribute something to 
our present knowledge of wave-lengths. In this connection 
also, especial attention will be paid to the fig of all 
measured distances in terms of the standard Meter of the 
Archives at a temperature convenient for use. With this view 
a standard decimeter Aen te into 10,000 equal parts will be 
taken as the unit of compariso 
As a type of the character of the work already done I will 
close this article with a description of two glass plates ruled on 
different days with the same setting of the stops of the magnet- 
arm. The stops were set to correspond to a motion of one 
thousandth of an inch upon the screw. The machine was 
started at 10 o’clock a. M. Between 12 and 1 o'clock it re- 
mained at rest. It was also at rest during the night. Starting 
again at 7 o'clock the next morning it was stopped at 9" 50", 
having ruled 4001 lines, covering a space of four inches. 
The plate was then removed, and another one was placed in 
position. The machine was again started, and this second plate 
was ruled under nearly the same conditions with regard to time 
and temperature as the first one. 
