AND MODERN MICROSCOPES. 137 



power of the objective that is important (for in many of the 

 trials here reported the lower powers have given the best 

 results, and the Tolles' ^ immersion the best on record), but 

 it is the skill of the optician in making the instrument. 



Mr. Greenleaf has just tried (February 7th) an immersion 

 objective by Wales' -',. He resolved the 10th, 11th, and 

 12th bands perfectly ; the loth was doubtful. Another trial 

 of the Hartnack No. 10 resolved the 13th band perfectly — 

 the 14th doubtfully. 



I have since tried the Wales' objective dnj, and resolved 

 the 13th band well, thus doing what Mr. G. did with it in 

 water ; the inference must be that Mr. G. did not obtain its 

 best work. 



Note. — Since the foregoing was written Dr. Barnard has 

 made more trials, and I am well satisfied that he has seen 

 the 19th band with a Spencer -pV and Tolles' i, both dry 

 objectives. This jJerformance fairly surpasses anything yet 

 done, either in this country or Europe. Dr. Barnard writes 

 (Jan. 29), that he found that the counting of the lines was 

 attended with the very difficulties referred to above, in addi- 

 tion to which there is another trouble, the whole width of a 

 band is not in perfect focus at once ; this necessitates a slight 

 change of focal adjustment, and any change renders it ex- 

 tremely difficult to fix, even with the cobweb micrometer, 

 the exact line last counted. He made five counts of the 

 19th band with the -^, namely — 



1. 110,392 to the English inch. 



2. 108,270 



3. 113,737 



4. 106,226 



5. 115,474 

 Mean, 110,820 



The number, according to Nobert, is 112,668. He counts 

 for the 15th, 91,545; Nobert, 90,074. Though there is 

 apparently considerable discrepancy between the count and 

 Nobert's figures, yet I consider it as near as can be reason- 

 ably expected when all the difficulties are appreciated. 

 Besides, it must be remembered that Dr. Barnard gives as 

 above the number of lines to an inch, not the number 

 actually counted. The actual number in the 19th band 

 should be 56*5, if the band is exactly -3-0V0 of an inch, a 

 variation of two lines each way covers the extremes of his 

 counting. 



