1882.] 



MICEOSCOPICAL JOUKNAL. 



25 



ing the divisions about two per cent, 

 too large, but the divisions are excep- 

 tionally even, as micrometer ruling 

 goes. The other has the average 

 length of the divisions nearly accu- 

 rate, but they are of inferior evenness. 



The power used in examining them 

 was about 450 diameters, as the 

 coarseness of the lines in the ordin- 

 ary micrometers was such that with 

 greater magnification the liability to 

 error in judging when the hair-line 

 of the eye-piece was in the middle of 

 the stage-micrometer line was more 

 than an offset to the advantages. 

 With Prof. Rogers' plate a higher 

 power could be made useful, because 

 the lines themselves are finer. 



The following is the result of a 

 series of average evenness of results 

 out of a number made with each mi- 

 crometer. 



Hundredths of ail inch — Plate No. i. 



Space— 12 3456789 



Measure--i5 44.8 45.3 45.4 45.1 45 45.2 45.4 45.1 



10 11 12 13 14 15 16 17 18 19 20 



45 45.3 45.5 45 45 45.2 45.4 45.2 45 45.1 45.5 



Plate No. 2. 



12 845 6789 10 



46 45.7 45.7 46 45.7 45.6 46 45.8 46 45.6 



Rogers' Plate. — The measurements 

 were 45 throughout twenty spaces 

 with no appreciable variation. This 

 was the more striking, because as the 

 ruling was to g-g^o^j-inch, the alternate 

 five and ten space long lines were 

 necessarily the boundaries of hun- 

 dredths, and there was no special rul- 

 ing of hundredth spaces on the plate. 



Thousandths of an inch — Plate No. i. 



1 23456789 10 



4.5 4.5 4.6 4.5 4.5 4.5 4.6 4.5 4.6 4.6 



Plate No. 2. 



1 23456 789 10 



4.6 4.5 4.5 4.6 4.7 4.75 4.6 4.7 4.6 4.6 



Rogers' Plate. — As there were no 

 spaces of .ooi-inch first, I took the least 

 number of spaces that would make 

 even divisions of the eye-piece 

 micrometer. As the test was only 

 for evenness, and not for absolute 

 measurement, I was not careful to 

 take the identical magnification, and 

 shortened the tube to accommodate 

 more readily the divisions of the eye- 



piece to those on the stage. I thus 

 made twenty divisions of the ocular, 

 equivalent to seven of the plate. 

 Twenty successive tests of this meas- 

 ure applied to twenty different groups 

 of seven spaces each, showed no vari- 

 ation whatever. The value of the 

 thousandth was thus made, .357 

 throughout. 



Next, I examined separately twenty 

 single spaces, each gg^oQ-inch, with a 

 higher power (one-tenth objective). 

 These were taken at irregular inter- 

 vals across the plate, and brought suc- 

 cessively to the same line in the eye- 

 piece, the tube being adjusted till 

 the first was as exactly as possible 

 equal to six spaces in the ocular. 

 The plate stood the last test quite as 

 well as before, each division precisely 

 filling the measure applied to it, and 

 being magnified 750 times. 



Of course it would not do to say 

 that there are absolutely no inequali- 

 ties in the spacing, but only that were 

 none measurable by the means em- 

 ployed, and that by whatever means 

 the divisions were measured, their 

 regularity and equality were some- 

 thing unparalleled in my experience. 



It is true also that the above gives 

 no test of the absolute accord of the 

 divisions with the standard inch or 

 centimetre. For that we must trust 

 Prof. Rogers' methods till some very 

 elaborate tests can be applied. But 

 it certainly shows both that the rela- 

 tion of the fractional inch to the cen- 

 timetre is exactly reproduced in the 

 plate, and that the subdivisions are 

 equal to each other within inappreci- 

 able limits. 



In the measurements I was of 

 course careful to use always the same 

 subdivisions of the eye-piece micro- 

 meter, so that no errors in its ruling 

 might mingle with the results. Simi- 

 lar tests applied to a plate ruled more 

 recently by Prof. Rogers produced 

 similar results, showing that the 

 acccuracy of the first is not excep- 

 tional, but is only a fair sample of the 

 ruling the Professor is able to do 

 with his new machine. 



