38 Tue Microscope. 
with so fine intervals the limits of accuracy at the eye end of 
the microscope are little, if any, less than the value of one half 
of one division. With the filar micrometer there is no such em- 
barrassment. The spider line can, when the definition is good, 
be brought into exact optical contact with the object to be 
measured, and the divisions read off without resort to estimation 
further than the fractions of one division of the index, and this 
_is so large as not to involve any difficulty. If the definition of 
the objective is bad, it is easier to use the filar than the glass 
eye-piece micrometer, but of course, exact measurements will in 
that case be impossible with any micrometer. When the am- 
plification is so increased by the use of high powers that one 
division of the ordinary glass micrometer represents a small 
fraction of a mikron, very good work may be done with sucha 
micrometer, as we think will be apparent by examinining the 
foregoing tables. 
With increased experience, my respect for low power ob- 
jectives used with the filar micrometer in making micrometric 
measurements and comparisons has greatly increased. When 
an object is large enough to occupy an appreciable portion of 
the field e. g. too inch, I am inclined to believe that as good re- 
sults in a series of measurements can be obtained with a low 
as with a moderately high power. At least such has been my 
experience. In measuring very minute objects or objects so 
small even as blood corpuscles, a power must of course be used 
that will give an amplification large enough to occupy an ap- 
preciable space in the field, but the instant the definition be- 
comes defective and the edges “wooly,” my experience is that 
better results can be obtained with a lower power that will give 
perfect definition. 
The conclusions from the above tests are obvious. When 
the greatest attainable accuracy is not required, the glass eye- 
piece micrometer is the more useful and convenient and is ac- 
curate at least up to the value of one division or even less, if 
the distance between its lines is greater than ,,55 inch, or a 
higher eye-piece than one inch is used. I have, however, never 
found it advisable to use a deeper eye-piece than one inch with 
any objective. 
To illustrate further: With a one inch eye-piece, tube- 
