250 



KNOWLEDGE. 



[November 1, 1895. 



and inaccuracies in the measurements of wave-length, etc. 

 Even with an ordinary wire grating, however, not at all 

 a perfect instrument, Fraunhofer in his early operations 

 on the solar spectrum obtained wonderfully accurate 

 measurements. 



In order to explain how a spectrum can be formed by 

 means of a diffraction grating, let us first consider the 

 phenomenon known as the interference of light, on which 

 the effect largely depends. If a light substance, such as 

 a piece of cork, was floating in water while a series of 

 waves was passing along, the cork would be observed to 

 move up and down nearly in a vertical line. The motion 

 of the water particles is like that of the cork, they move 

 up and down in straight lines. Now if instead of one set 

 of waves we had two, matters might be so arranged that 

 the crests of one set of waves fell on the hollows of the 

 other set ; when this took place we should have what is 

 called interference of the waves. The water surface, 

 when this interference took place, would not be disturbed 

 if the two wave-systems were of equal magnitude ; no 

 waves would be seen ; and the cork would be at rest on 

 the surface. One wave-system would urge it upwards and 

 the other downwards, and these forces on it would just 

 balance each other when the two wave-systems were 

 present. Interference also occurs when the crests of one 

 pystem of waves fall on the crests of another ; the 

 amplitude of the wave disturbance is increased, that is, 

 the distance from the bottom of the trough to the top of 

 the crest is made greater. 



In the case of the grating, the diffraction and interference 

 of light produce effects which resemble those of refraction 

 in the case of the prism. To explain how this occurs let 

 us refer to the diagram, in which A B C D represent four 

 openings in a grating, or four clear spaces between the 

 rulings. Let us suppose the front of the advancing light- 

 wave is parallel to the grating, and that the diagram 

 represents a section through the grating at right angles to 

 the plane of the grating ; then we may consider ABC 

 and D to be four points at equal distances apart, emitting 

 light of the same wave-length, the vibrations being such 

 that a crest starts from each of the points A B C D at the 

 same time. Consider the light which travels from 

 A B C D in the direction D L, and let a lens L be 

 interposed in its path. Draw lines D M, C N, B P, at 

 right angles to the direction D L of the light. These lines 

 will cut off lengths K C, R Q, Q B, M N, N P, P A from 

 the lines D L, etc., representing the rays, and these lengths 

 are from the geometry of the figure all equal. If the 

 lengths A P, P N, etc., be taken to be the magnification of 

 a wave-length of light, when a crest of a wave is at A 

 crests will also be at P, N, and M, and midway between 

 these points will be the troughs of the waves. Similarly, 

 at the same moment, crests will occupy the positions Q, 

 R, and K. The vibrations, then, are in the same phase 

 along the line ]> K R M, crests at the same instant being 

 at D, K, etc. The interposition of the lens L causes the 

 rays to come to a point at F, the focus of the lens, and the 

 time taken by the light to travel from D, K, E, and M to 

 F is the Eame ; therefore, the waves being in the same 

 phase along the line D M, they will be in the same phase 

 at F — that is, the crests will all coincide at F, and that 

 point will be brightly lit up. Light of a different wave- 

 length will be brought to a focus different from F, but near 

 to it, and the effect will be that of a number of points or 

 lines, each consisting of illumination of a different wave- 

 length and different colour placed close together — that is, 

 a spectrum is formed. Taking a similar figure in which 

 the distances C K, Q R, etc., represent half wave-lengths, 

 when a crest is at A a trough of a wave will be at P, 



another crest at N, and so on ; and when a crest is situated 

 at D a trough wiU be passing over K, a crest over R, and 

 another trough over M. Thus at F crests and hollows 

 will be brought together in equal number, and the result 

 will be absence of illumination at that point — F will be 

 dark. The position of F for light of a certain wave-length 

 will depend on the magnitude of that wave-length ; for 

 light of a slightly differing wave-length, F will be situated 

 at another point, but close to the first. Thus, considering 

 A, B, C and D a section through a grating, the effect on 

 a screen passing thi-ough F, or the appearance to an eye 

 looking through a telescope of which L is the object-glass, 

 will be that of a band of light, each line in which is 

 produced by light of different wave-length, and therefore 

 of differing colour, these lines being distinct and not 

 overlapping. This result is expressed by saying that a 

 pure spectrum is produced. A pure spectrum is one in 

 which overlapping of light of different wave-length and 

 colour does not take place. 



A great advantage which a spectrum formed by a grating 

 has over that given by a prism proceeds from the fact that 

 the dispersion depends only on the wave-length and the 

 closeness of the rulings. Thus any one diffraction spec- 

 trum is an exact copy of any other on a larger or smaller 

 scale. The ratio of the widths occupied in the spectrum 

 by any two colours is the same in all diffraction spectra. 

 This does not hold in the case of the spectra formed by 

 prisms. The relative dispersion in spreading out of the 

 colours differs with different prisms according to the kinds 

 of glass of which they are made. Thus prism spectra can- 

 not be compared together, but observations with a grating 

 can be compared with others wherever taken and with 

 whatever instrument. On this account the diffraction 

 spectrum is chosen as the standard. 



(Tn he contifiued.J 



&titmt NotC9. 



The following interesting results of experiments relating 

 to the growth of trees at different times of the day have 

 been sent to us by Mr. E. H. Thompson, the Government 

 entomologist of Tasmania. Measurements were taken as far 

 as possible every three hours, with the following results: — 



From 6 a.m. to 9 a m 8| per cent, of gi-owtli 



,, 9 a.m. to noon ... Ij ,, ,, 



,, iioou to 3 p m. , , No growth 



,, 3 p.m. to p.m. ... ,, 



., (5 p.m. to 9 p.m. . ... li per cent, of growth 



,, 9 p.m. to lii p.m. ... ... 3J ,, ,, 



12 p.m. to 6 a.m. ... ... 8.3 ,, ., 



The greatest growths in twenty-four hours were banksia 

 rose, six and a half inches; geranium, five and three-quarter 

 inches ; wattle, four and one-third inches ; apple, two and 

 a quarter inches ; pear, one and a third inches. 



Writing from Lippentott, Manitoba, Mr. Arthur Parry 

 says that the sun pillar, described by Mr. Barber in the 

 •Tune number of KNowLEDfiE, has been more than once 

 visible there during the past winter, while on Jan. 10th 

 and 30th a similar shaft was visible proceeding from 

 the moon. With regard to two other phenomena which 

 Mr. Parry would like explained, he writes, on Jan. 12th, 

 a " circle round the moon, with curious shaft of light 

 obliquely across it, visible at 11 p.m. for a short time " ; 

 and on March 9th, " bright moondogs and circle above. 

 The peculiarity of which lies in the fact that the circle 

 which should have connected the moondogs was not visible, 

 while the circle which had no visible connection with 

 them was bright and complete." 



