814 SUMMARY OF CURRENT RESEARCHES RELATING TO 



(5) Microscopical Optics and Manipiilation. 



Variation in Micrometric Measurements due to different illumi- 

 nation.— Mr, C. Fftsoldt sends us the following " Tabic showing tlio 

 variation in measurements due to the ditTerent applications of light and 

 illuminations." 



" Tho imafTc of 4/10 in. was the object on which these measure- 

 ments were made, and was ruled on a glass disc of No. 2 covering glass, 

 7/1000 in. in thickness. 



" All mcasnromeuts were taken on one and the same ruling, with the 

 same Microscope, objective, and eye-piece, under the same focus, and 

 liavinfT the Microscope in the same position continually, and only 

 changing tho mirror and excluding the one light while the other was 

 used. 



Unmoun ted — Lamplijh t. 



Lines downwarj. 

 Concave mirror 4/10 in. 10/100,000 - 

 Plane „ 4/10 in. 5/100,000 + 

 111. through I ^ joi,^ 5/100,000 + 

 objective J 



Lines upward. 



Concave mirror 4/10 in. 10/100,000 + 



riane „ 4/10 in. 14/100,000 + 



1. tlirongh I 4/10 in. 15/100,000 + 

 objective ) ' / > ■ 



111. through 

 ohje 



Mounted on Glass 



Daylight. 

 Concave mirror 4/10 in. 30/100,000 + 

 Plane „ 4/10 in. 20/100,000 + 



Lamplight. 

 Concave mirror 4/10 in. 

 Plane „ 4/10 in. 15/100,000 + 

 111. through I wjQ .jj 31/100,000 + 

 objective j ' 



" A number of comparisons were made at each position ami in the 

 same temperature. 



" A Spencer objective was used for these measurements ; but Bausch 

 and Lomb and Guudlach objectives were also tried, obtaining the same 

 results. 



" The Microscope used is one constructed on my late patents, and 

 has a micrometer for measuring similar to a cobweb micrometer. Bnt 

 instead of cobwebs, thi'ee movable steel pointers are used, which arc 

 worked as fine as this metal will permit. The stage is mechanical, and 

 the main slide is moved with great precision by a fine screw 100 threads 

 per inch." 



Error was therefore eliminated in the case only of the lines mounted 

 on glass when the concave mirror and lamplight was used. 



Testing Screw-Micrometers of Eeading-Mieroscopes.* — Prof. Eein- 

 hertz points out that every micrometer is liable to special errors and 

 that these must be studied before the requisite corrections can be 

 applied. The errors are due to (1) the screw itself; (2) the mounting of 

 the screw ; (3) the remaining parts of the micrometer. 



(1) According as the screw 2)roduccs unequal linear movements at 

 different parts for a complete turn, or unequal linear movements for 

 equal fractions of a single turn, the errors may be called " jirogressive " 

 or " periodic " ; the former are due to inequalities of pitcli, the latter to 

 irregularities of the thread. 



(2) The position of the screw is fixed by its point or head being 

 maintained in constant pressure against a plane surface ; if this surface 



* Ccutral-Ztg. f. Optik u. iMech., ix. (1888) pp. 37-40. 



