410 



SCIENCE. 



[N. S. Vol. XIV. No. 350. 



of the prism, gave the data for plotting the dis- 

 persion curve of rock-salt. The results, ex- 

 ceeding in accuracy any heretofore obtained, 

 and which will be of the greatest value to other 

 investigators using rock-salt dispersion for infra- 

 red work — are given in several convenient 

 forms, most conveniently, perhaps, in the shape 

 of a very large scale curve, extending between 

 the limits ^ = 0.5 /J- and ^ = 6.5 z^, from which 

 indices of refraction can be read off to 0.000002, 

 and for which the probable error lies usually 

 between 0.000009 and .000018. The dispersion 

 curve of fluorite is given on the same scale 

 between the limits A = .75 ,« and /^ = 3.5 /x. 

 From the data for rock-salt have been calcu- 

 lated the five constants of the Ketteler disper- 

 sion formula, which differ quite noticeably from 

 those calculated by Rubens and Trowbridge for 

 a longer wave-length interval ; but the differ- 

 ences between the observed and computed 

 values of index of refraction are hardly greater 

 than the probable errors of deviation, except 

 for the longest wave-lengths. Accurate com- 

 parisons have also been made of the dispersion 

 of three rock salt prisms which confirm the view 

 that rock-salt as found in nature is in one re- 

 spect of great optical uniformity, so that accu- 

 rate determinations of indices of refraction for 

 one prism can be safely applied to another for 

 the purpose of determining wave-lengths ; an 

 extremely fortunate circumstance. 



It is impossible to do more than mention 

 some of the lesser pieces of work here recorded 

 — such as the determination of the energy 

 curves of various incandescent mantles; tests 

 of the accuracy or constancy of the bolometer, 

 which unfortunately do not touch the most 

 diflScult point, i. e., constancy as regards sensi- 

 bility for long time intervals ; and a study of 

 the minute structure of the infra red absorption 

 band 'wj.' A considerable space is, appro- 

 priately, devoted to the discussion of errors, and 

 to methods for overcoming diflBiculties inherent 

 in bolometric work ; of which the most trouble- 

 some are undoubtedly 'drift,' or the continu- 

 ous change in the zero position of the galvan- 

 ometer spot of light, and the more or less rapid 

 periodic changes of zero, or 'wiggle.' The 

 various precautions taken, which have finally 

 almost eliminated the drift and greatly reduced 



the * wiggle, ' are dealt with in full ; few of 

 them altogether new perhaps, but certainly ap- 

 plied here with greater completeness and care 

 than has been done elsewhere. The detailed 

 consideration of manipulation and construction 

 will be of great value to others engaged in 

 similar work. In particular should be men- 

 tioned the study of the behavior of rock-salt 

 prisms under conditions of rising temperature ; 

 the question of the construction of linear bo- 

 lometers, in which such skill has been attained, 

 and of the design of balancing bridges ; and the 

 full discussion of the adjustments of the fixed- 

 arm spectroscope. In the chapter on the gal- 

 vanometer will be found a useful table of the 

 computed axial magnetic force produced by 

 galvanometer coils of various resistances, 

 wound either .with a single size of wire or with 

 different combinations of three sizes, which not 

 only shows clearly the advantage of sectional 

 winding, but will be a valuable aid in the de- 

 sign of galvanometers. The use of the Ayrton- 

 Mather scale for expressing galvanometer sen- 

 sibilities, now so generally adopted, would have 

 rendered easier the comparison of the observa- 

 tory instrument with others. 



The valuable and interesting material, which 

 it has been here attempted briefly to sum- 

 marize, is the result of about nine years of 

 work, involving the labors, successively, of 

 Dr. Hallock, Mr. Wadsworth, Mr. Child and, 

 for a longer time, of Mr. Abbot and Mr. Fowle, 

 all under the direction of Professor Langley. 

 From time to time statements of the progress 

 of the work have been made by Professor 

 Langley, and some special points considered in 

 papers of Mr. Wadsworth, Mr. Abbot and Mr. 

 Fowle, but the results now made public are so 

 interesting and so important to all engaged in 

 similar work that it is to be hoped that in the 

 future conditions affecting publication by the 

 government may allow more frequent and less 

 delayed reports from this observatory, which 

 is unique in its possibilities for the highest 

 class of work with radiant energy. 



C. E. Mendenhall. 



Dynamo Electric Machinery. Its construction, 

 design, and operation. Direct current ma- 

 chines. By Samuel Sheldon, A.M., Ph.D. 



