ZOOLOGY AND BOTANY, MICROSCOPY, ETC. 447 



subject be thought of sufficient importance and interest, I will describe 

 the process by which any good observer, who is the owner of a filar 

 micrometer, and who knows the correction for total length of his micro- 

 meter, and last but not least, who has sufficient patience, can determine 

 the errors of any subdivisions small enough to be brought within the 

 field of his Microscope." 



Micrometer Measurements.* — Dr. M. D. Ewell, referring to his 

 advocacy of the use of metal micrometers uncovered, recommends that if 

 such a temporary cover should be used, it should always be used under 

 precisely the same conditions, and the observer should be quite sure that 

 both faces of such cover are jiarallel, otherwise the influence of refrac- 

 tion, the cover acting as a prism at some part of its surface, might 

 introduce errors of unknown magnitude. For this reason, on further 

 reflection, he thinks it better to have a permanent cover on micrometers 

 intended for use with high power objectives, and to have the corrections 

 of such micrometer determined with such cover in situ. 



This leads him to notice a table of measurements published by 

 Mr. C. Fasoldt.t intended to invalidate the result of the investigation 

 of Centimetre Scale " A " of the American Society of Microscopists, 

 and its so-called copies, by the different observers who have investigated 

 them. As to this Dr. Ewell says : — " Mr. Fasoldt does not in his pub- 

 lished paper give sufficient data to enable one intelligently to criticize or 

 judge of the accuracy of his work ; but there is one element of uncer- 

 tainty about it that seems quite patent, viz. that it does not appear that 

 the glass disc upon which the lines were ruled had either sui'face plane, 

 or that the two surfaces of the disc were parallel. If nothing else 

 appeared, to my mind the fact that the space was measured with different 

 sorts of illumination, and with the lines first downward and then 

 upward, thus introducing unknown errors due to the causes above 

 specified, would deprive the results of any value they might otherwise 

 possess. There is no means of intercomparison and of eliminating 

 these unknown errors. 



I cannot ascertain, however, from the paper, with what standard the 

 4/10 in. was compared, or exactly how it was compared. If, as 1 

 suppose, it was compared with the screw of a screw stage-micrometer, 

 which was assiuned to be a constant, I must beg to dissent from any 

 conclusion thus obtained. I find it necessary, in ruling standards of any 

 considerable length, to assume a value for the screw, rule a trial scale, 

 and by actual comparison with some authentic standard deduce therefrom 

 a series of corrections before ruling the final scale. If great accuracy 

 is desired, it may be necessary to repeat this several times before ruling 

 the final scale ; and this is the case notwithstanding the errors of the 

 screw have previously been carefully investigated. I would never trust 

 any screw or train of wheels as a final standard of reference for more 

 than about one-half the field of the Microscope, much less for so long a 

 space as 4/10 in." 



Klaatsch's Radial Micrometer.^ — The radial micrometer of Dr. H. 

 Klaatsch consists of an eye-piece micrometer-disc, not only subdivided 

 along the usual straight line, but traversed by two diameter lines cutting 

 each other at right angles, and both of which are subdivided. In two 



* The Microscope, ix. (1889) pp. 74-6. 



t See this Journal, 1888, p. 814. j Auat, Auzeig., ii, (1887) pp. G32-4. 



