July 9, 1915] 



SCIENCE 



59 



most serious one, it was readily detected by 

 Dr. Eavn, and after calling it to the attention 

 of others it was not at all difficult to recognize 

 it again. 



It has always been a matter of surprise to 

 the writer that this rust has not occurred in 

 North America before, the rust being so com- 

 mon in Europe and samples of wheat con- 

 stantly passing back and forth. Nevertheless, 

 its existence this season in such abundance in 

 portions of Oregon and Washington makes it 

 evident that either the rust has increased with 

 remarkable rapidity or has already existed in 

 the country for several years. The latter, if 

 true, would be in face of the fact that it is 

 easily distinguished from other rusts and that 

 pathologists have been actively studying the 

 rusts of the country. Further details of the 

 occurrence of the rust will be reported later. 

 M. A. Carleton 



July 3, 1915 



SCIENTIFIC BOOKS 

 A Study of the Orbits of Eclipsing Binaries. 



By Harlow Shaplet. Contributions from 



the Princeton University Observatory, No. 



3, 1915. 4to. Pp. vii + lY6. 



In astronomieal literature one of the most 

 frequent subjects refers to the "orbit" of a 

 heavenly body. In fact, for a long time a 

 standard topic for a doctor's thesis was the 

 determination of the definitive orbit of a 

 comet. Here the task of the candidate was to 

 derive from observations, made in all parts of 

 the world, the best possible numerical values 

 for the six elements or constants which de- 

 fine the path of the comet as a conic section in 

 space with the sun at the focus. Other classes 

 of orbits are those of visual double stars, 

 spectroscopic binaries, and finally, as in the 

 work under review, we have what may be 

 called photometric orbits, since the results are 

 based upon observations of the light variations 

 of stars. 



Even to those familiar with the subject, the 

 amount of mathematical analysis that has 

 been based upon the changes of some of the 

 variable stars is a source of wonder. As an 

 illustration may be mentioned the famous 

 star Algol, which has been the subject of half 



a dozen extended monographs, scores of papers 

 and literally tens of thousands of observations. 

 The special importance of the stars whose 

 variations are due to the eclipses of large close 

 companions is due to the fact that these sys- 

 tems give us the only satisfactory clue to the 

 actual diameters of stellar bodies. The theory 

 of such cases has been well understood for a 

 long time, but recently Professor H. N. Rus- 

 sell, of Princeton University, has developed a 

 new method for determining the elements of 

 eclipsing binaries. He recognizes the fact that 

 measures of the light of stars are seldom if 

 ever accurate to one per cent., so that approxi- 

 mate and graphical methods are sufficient for 

 any case that can arise. In essence his method 

 consists of solving not directly for the ele- 

 ments of a double system, but for the best 

 light-curve that will represent the observa- 

 tions, and then the characteristics of the 

 system are easily computed from the curve. 

 A series of papers on this general subject have 

 appeared by EusseU and Shapley, and the pres- 

 ent contribution summarizes much of the 

 previous work. Though not so stated, it is 

 understood that this is a thesis for the doctor's 

 degree, with subsequent additions to bring the 

 work into complete form. 



Whereas formerly an exhaustive study of 

 one star was thought to be quite a piece of 

 work. Dr. Shapley with the new methods has 

 undertaken and carried through a pretty thor- 

 ough discussion of 90 eclipsing stars, or all for 

 which any sufficient data exist at the present 

 time. We learn that the discussion of a single 

 object required not less than a day, nor more 

 than two weeks. Even though nearly all of 

 the observational material was already avail- 

 able, it was a considerable task of mere rou- 

 tine to get it together, and one of the advan- 

 tages of this memoir is that it will serve as an 

 index to the best sources of information con- 

 cerning any particular star. 



There is a vast difference in the quality and 

 completeness of the data for different systems, 

 and many of the numerical results are avow- 

 edly only rough, or perhaps even guesses. In 

 particular, the proportion of light which comes 

 from the fainter component, as indicated by 

 the secondary eclipse, has to be assumed in 



