C. S. Petrce—Wave-length Comparison. 51 
Art. VI.—WNote on the Progress of Experiments for comparing a 
Wave-length with a Meter ; by C. S. Petrcz. “Communicated 
by the Superintendent of the U. S. Coast and Geodetic 
Survey. | 
To C. P. Patrerson, Superintendent U. S. Coast and Geodetic Survey :— : 
Dear Sir—The following is the present state of the spec- 
trum meter business. The deviation of a spectral line (Van 
der Willigen’s No. 16) has had three complete measures using 
a certain gitter of 3403 lines to the millimeter. The double 
deviation (the angle measured) was found to be 
1877. June 23, 89° 54’ 1975 
June 29 and July 2, 19°25 
Sep. 4 and Aug, 27, 19°65 
Mean 89° 54’ 1975 
An error of 0:4” in this would occasion an error of one mi- 
cron in the meter. These measures were previously commu- 
nicated to you, but owing to an erroneous value of the coeffi- 
cient of expansion of glass having been used (the value for 
iron having been inadvertently substituted) they did not seem 
to agree as well as they do. There were two other complete 
measures, but in regard to one of them there is a doubt about 
the thermometer used, and in regard to the other there isa 
doubt about the part of the line set on. This line seems on the 
whole to be a bad one for the purpose. Another line near it 
was therefore selected and another much finer gitter. The 
deviations obtained were on the different days: 
1879. May 8, 90° 03’ 517-7 May 15, 90° 03! 50°35 
May 9, 1°75 | May 21, 51°75 
May 10, y 22, 51-2 
Notwithstanding the bad result of May 15, which is unaccourt- 
abie, these measures are evidently good enough. One of these 
gitters has been compared with all the centimeters of a deci- 
meter scale of centimeters. The other is still to be compared 
with all the even two centimeters of the same scale. : 
r. Chapman is now comparing this decimeter scale with all 
the decimeters of a meter scale of decimeters. As soon as that 
is done a meter will have been compared with a wave-length. 
But shortly after, this will be improved by comparing the other 
gitter and also a third -which I propose to measure a 
deviation. It will remain, first, to find the coefficient of expan- 
sion of the glass meter. The apparatus is all ready for this and 
it will-not take a fortnight. Second, the glass meter will have 
be compared with a brass meter. This will be an operation 
of some difficulty but I think we shall ag oct it before long. 
Yours respectfully, C. 8S. Perrce, Assistani. 
’ 
