PHYSICS. 605 



inches ; hence the spectrum thrown by it on a flat plate is normal within 

 about 1 part in 1.000,000 for 6 inches, and less than 1 part in 35,000 for 

 18 inches. In photographing the spectrum on a flat plate the definition 

 is excellent for 12 inches, and by the use of a plate bent to 11 feet ra- 

 dius a plate of 20 inches in length is in perfect focus, the spectrum 

 being so nearly normal that for most purposes its error may be neglected. 

 Another important property of the concave grating is that all the sup- 

 erimposed spectra are in focus at the same point; so that the relative 

 wave-lengths are readily determined by micrometric measurement. 

 Knowing, therefore, the absolute wave-length of one line, the entire 

 spectrum can be measured. This method is the most accurate known, 

 as by simple inspection the relative wave-length can be judged of to 1 

 part in 20,000, and with a micrometer to 1 part in 1,000,000. This 

 method 'is especially valuable in obtaining the focus in the invisible 

 parts of the spectrum. Examining the question whether the ruling ac- 

 tually performed, in which equal spaces are ruled along the chord, could 

 be replaced to advantage by any other kind of ruling, the author finds 

 that the departure of the ruling from theoretical perfection is of little 

 consequence until lines twenty times as fine as the 1474 line can be 

 divided ; the components of this line being one forty -thousandth of the 

 wave-length apart. Considering, finally, the question of the limit of the 

 resolving power of the spectroscope, he shows that all lines have some 

 physical width and that we are limited by that width in the resolving 

 power of the spectroscope. All the methods of determining the limits 

 seem to point to about the 150,000th of the wave-length as the smallest 

 distance at which the two lines can be separated in the solar spectrum 

 by a spectroscope of even an infinite power. Practically he has been 

 able to photograph bines which do not differ in wave-length more than 

 one part in 80,000, and he believes he can resolve lines whose compo- 

 nents are only one 100,000th of the wave-length apart. So that the 

 idea of a limit has not yet been proved. (Am. J. Sci., August, 1883, 

 m, xxvi, 87.) 



Similar investigations on the theory of concave gratings have been 

 made by Mascart (J. Phys., January, 1883, II, II, 5), by Baily (Phil. 

 Mag., March, 1883, V, xv, 183), and by Glazebrook (Phil. Mag., June, 

 1883, V, xv, 414). In a note, subsequent to bis last paper, Eowland 

 has called attention to certain errors in the latter paper, intimating 

 that some of the methods suggested were identical with those he had 

 himself presented to the London Physical Society six months before. 

 Indeed, in a foot-note to his previous paper, he had expressed his sur- 

 prise at this invasion of his field by others, saying that he had expected 

 to be allowed a little time to work up the subject himself. (Am. J. Sci., 

 September, 1883, III, xxvi, 214.) 



Glazebrook has suggested a new form of polarizing prism, free from 

 the defect of the Nicol, of displacing laterally the object seen through it. 

 It is made by cutting a rectangular parallelopiped from a piece of spar 



