September 4, 1885.] 



SCIENCE. 



183 



not be of immediate necessity in the conser- 

 vative lines of astronomical inqairy. 



David P. Todd. 



LETTERS TO THE EDITOR. 



%* Correspondents are requested to be as brief as possible. The 

 writer's name is in all cases required as proof of good faith. 



A method for determining the unit of light. 



In all photometric work hitherto undertaken, 

 one of the main difficulties has been to obtain a 

 satisfactory standard of light, — one which will be 

 always constant, and which can be accurately dupli- 

 cated. Heretofore, all experiments in this direction 

 have been failures. The plan here suggested con- 

 templates, not the employment of a unit quantity of 

 light, but the employment of a certain effect pro- 

 duced by that unit quantity as a standard. In other 

 words, it makes not the light, but the photometer, 

 the constant. 



This photometer must, then, be some device for 

 measuring radiant energy. But, for photometric pur- 

 poses, we wish only to measure that portion of the 

 energy which has a wave-length readily visible to 

 the human eye. With the great differences in color 

 of our modern sources of illumination, it is absolutely 

 impossible to state an exact equivalence between the 

 yellow light of a candle-flame, and the blue light of 

 an electric arc. For really accurate work, we can 

 compare only light rays of the same wave-length. 

 As the human eye is most sensitive to light from 

 that portion of the spectrum between the D and E 

 lines, in the following plan I have selected that re- 

 gion of the spectrum to be used exclusively for the 

 comparison of the brilliancy of the various lights. 

 In all probability, the total brilliancy of an incandes- 

 cent body does not increase in a ratio exactly pro- 

 portional to the increase in brilliancy of the yellow 

 rays; but this difference, within practicable limits, is 

 probably so small as to be entirely negligable. And 

 we have the advantage of being able to state an ac- 

 curate arithmetical ratio between the lights, instead 

 of what must be at best a mere general comparison of 

 the relative effect of the two lights upon our eyes. 



Briefly stated, tben, the method I would suggest 

 consists in moving the light to be measured towards 

 the slit of a spectroscope, until a certain effect is pro- 

 duced upon a screen so placed as to receive the 

 yellow rays. When this effect is produced, the spec- 

 troscope is receiving the standard amount of light 

 from the source; and the brilliancy of the source can 

 then be determined by measuring its distance from 

 the slit. 



In attempting to apply this method, the difiiculty 

 which at first arises is, to obtain an effect which can 

 be measured with accuracy. By permitting the spec- 

 trum of a light to fall upon suitable screens, three 

 classes of effects may be obtained; namely, heat- 

 ing, visual, and chemical. Of these, the second is 

 evidently unsuited for the purpose of obtaining a 

 standard. The third is too uncertain, and not sus- 

 ceptible of sufficient accuracy, so that the first alone 

 remains. Of the two practicable heat methods of 

 measuring radiant energy, the thermopile is the more 

 sensitive; but the bolometer responds the quicker to 

 changes of temperature, and has the narrower sur- 

 face. The latter instrument has, therefore, been 

 selected for this application of the method. The 

 unexposed arm of the bolometer has a sHght addi- 

 tional adjustable resistance thrown into its circuit, 

 80 that, when the instrument is not in use, the 



needle of the galvonometer will have a certain deflec- 

 tion dependent on the strength of the battery-cur- 

 rent employed. When the light to be measured is 

 placed in front of the slit of the spectroscope (which 

 should be quite broad), the deflection will be dimin- 

 ished. As the light approaches the slit, the deflec- 

 tion will decrease, and finally become zero, at which 

 time it is giving out the standard light. Its bril- 

 liancy can now be read off from its position upon a 

 scale placed in front of the slit and parallel to the 

 collimator. 



This photometer might also be used to adjust the 

 position at which an incandescent electric or other 

 lamp should be placed in order to furnish a constant 

 supply of light. This source could then be used as a 

 unit in an ordinary photometer. 



Wm. H. Pickekixg. 



An American Silurian scorpion. 



The 'American scorpion' from the water-lime 

 group of New- York State, described by Professor 

 Whitfield on pp. 87, 88, is undoubtedly a young speci- 

 men of Eusarcus scorpionis (Grote and Pitt: Bul- 

 letin of the Buffalo society of natural sciences, vol. 

 iii., pp. 1, 2), so named by an error, and which will 

 be redescribed as Eurypterus scorpionis in the forth- 

 coming vol. V. of the society's bulletin. 



The enclosed is a sketch of the youngest specimen 

 in my possession, drawn full size: the largest I have. 



indicates an animal at least three feet long. There 

 cannot be any doubt as to its zoological position ; for 

 the characteristics of the genus Eurypterus — eyes 

 placed within the margin of the carapace, and a tri- 

 angular spine as caudal appendage — can be distinct- 

 ly identified. 



All my specimens have been found in the water- 

 lime group at Buffalo, associated with Eurypterus. 

 Pterygotus, and Ceratiocaris. Julius Pohlmax. 



Buflfalo, N.T., Aug. 5. 



