UNIFORMITY OF NOMENCLATURE IN OBJECTIVES. 139 
the opticians and is not free from question as to what standard is 
meant after all. The lens made as a standard is probably not a 
one-inch lens at all (principal focus), for the principal focus is never 
used in the microscope; and authorities differ as to whether it 
should have conjugate foci of one-inch and ten inches, or ten inches 
apart (one-inch and nine inches), Assuming x 10 as a one-inch 
power, would be most easily applicable and unmistakable; and 
_ this power, ten, divided by the ascertained power of any ocular 
or objective would give the equivalent focal length of that objective 
or ocular without comparison and beyond dispute. 
The chances -of error in this case are the same as in ordinary 
micrometry, with one or two additions, and should in all cases be 
ascertained in order to test the reliability of any series of observa- 
ions. They are due to the uncertain value of the divisions of the 
stage micrometer, to the like (but less important) variability of 
the measuring scale, to the uncertainty as to the exact optical cor- 
respondence of the lines selected for comparison in the two scales, 
and to the uncertainty as to obtaining exactly the assumed dis- 
tance between the upper scale and a given point of the objective. 
The first of these errors is the largest, and its magnitude would 
surprise many who have noticed and admired the remarkable 
“perfection” of the common micrometers. A micrometer which 
ought to be the best in the writer’s possession, with lines 100, 1000 
and 2000 to the inch, has a certain error of .02 and a limit of error 
of .035. This is entirely too much latitude for a single source of 
error, and of course it is nearly eliminated by comparing a large 
number of spaces belonging to at least several different scales, 
rejecting any scales which by differing widely from the average 
standard are presumably erroneous, and averaging the rest. The 
remaining sources of error may be similarly reduced by averaging, 
though their aggregate limit of error, ascertained by comparing the 
average measurements with extreme figures beyond which there is 
va Possibility of doubt, will be found to be very small and incon- - 
siderable. 
The standard distance of measurement in estimating magnifying 
powers may be stated to be, at present, ten inches. The distance 
of five inches has been recommended, even somewhat recently,* 
and eight,+ nine,t and ten§ inches have been successively used. 
* Brocklesby, N, Y., 1851. + Baker, Lond., 1742. 
t Fokes, Esq.. P.R.S., 1742. § Lardner, Carpenter, Suffolk, etc. 
