Outerbridge.] -1-04: [May 15, 



presented themselves in the outset, and it was some time before I succeed- 

 ed in obtaining a special adjustment of the apparatus appropriate to my 

 purpose. 



Using one-half of the largest Kitchie induction coil, throwing a spark 

 of eleven inches, (fed by a powerful battery and reinforced by four large 

 condensers) in connection with a two-prism Browning Spectroscope, I 

 found that upon gradually separating the metallic electrodes, certain of 

 the lines broke in the middle ; and, upon further increasing the distance 

 between the electrodes, the hiatuses in the spectral lines increased pro- 

 portionately, hut unequally with different alloys. 



This, as I am informed, is the novelty in spectroscopic research, dis- 

 covered by Mr. Lockyer, upon which the theory of possible quantitative 

 analysis is founded, and I was much gratified at having verified the ex- 

 periment. 



Repeated trials with various alloys, gave similar effects. Having proved 

 this general incident, a systematic series of experiments with alloys 

 enabled me to map the difference of fineness between specimens 500 and 

 750 fine and even to recognize the variation between ingot-slips 895 and 

 903 fine. These results were observed by Mr. Andrew Mason, of the New 

 York Assay Office, and by several members of the National Academy of 

 Sciences, then on a visit to the Stevens Institute, as also by other 

 gentlemen, to whom some of the experiments were shown. The varia- 

 tion within seven thousandths above referred to, was by no means 

 marked — indeed, over-cautiousness prevents my relying upon its certainty 

 — although a more delicate adjustment of apparatus and farther expe- 

 rience would probably render the distinction more decided. Of course, 

 in these experiments, it was necessary to eliminate the numerous air 

 lines which appeared in all the spectra. A difficulty which presented 

 itself in the exact comparison of certain characteristic lines of gold, 

 silver and copper, whose positions in the spectrum are in close proximity, 

 was overcome by using a pure metal as one electrode and another pure 

 metal as the other electrode. The effect thereby produced was very 

 curious. With pure gold and pure copper as the electrodes, the gold lines 

 extend across only one-half the field of the spectrum, and the copper lines 

 extend across only the other half, the medial termini of both sets of lines 

 being perfectly sharp and bright. By this means a double spectrum of 

 copper and gold is obtained, or rather, a section of a complete gold spec- 

 trum and a section of a complete copper spectrum are visible in imme- 

 diate juxtaposition, thereby enabling a most accurate comparison of lines, 

 which in reality are not identical in position, but which by the pre- 

 vious method were apparently so. 



By a slight modification of the experiment, substituting pure copper 

 as one electrode and an alloy of silver and gold as the other, the proxi- 

 mate lines of these three metals are presented mapped, as it were, on a 



