Mitchell^ on the Cobiveb Micrometer. 
73 
It has been said^ not only that the measurements afforded 
by the cobweb micrometer are unnecessarily delicate^ but, 
which is a contradiction, that they are also not so exact as 
they appear to be, and that Jackson^s glass micrometer is 
sufficiently accurate for all purposes. To this I beg to reply, 
first, that measurements, to be of any scientific value, should 
be made with the most perfect instrument which science and 
manufacturing skill have placed at our disposal. Secondly, 
that any one who will take the trouble to obtain the mean of 
a number of carefully made measurements of the best stage 
micrometer he can procure can always get a reading, cer- 
tainly and easily, within one division of his micrometer ; I can 
always do it within half that quantity, when I desire to be 
very exact ; in this method nothing is obtained by estima- 
tion, it is a simple mechanical operation. Thirdly, in Jackson^s 
glass micrometer the divisions are fixed, and no object is 
measured exactly that does not exactly reach from one division 
to the other. Such cases form the exceptions, and in the 
majority of cases there is something left to be estimated, 
which means simply guessed at. There should be nothing 
left to guess that can be measured, as it undoubtedly can 
with the cobweb micrometer. I have heard of such guessing 
of the diameter of blood-corpuscles when a man's life was iu 
the balance, 
