36 CELESTIAL, ATMOSPHEEIC, AND 



Sun.* It is scarcely necessary to say that the streaks of 

 light producing colour effects at sunset and sunrise are often 

 seen above and below the Sun, as well as to the right and 

 left. He seems to have believed that, unlike parhelia, the 

 " rods " were caused by reflection of a cloud, probably a 

 white one, in certain very watery clouds near the Sun. In 

 order to explain his meaning, he refers to the appearance of 

 a cloud when seen directly and when seen by reflection in 

 water. In the former case, he says, the cloud is quite 

 colourless, but, when seen in water, it seems to be full of 

 " rods."t Again, in an earlier passage, he says that the 

 reflection of the cloud in water is some colom* of the rain- 

 bow, for the visual rays being weakened in consequence of 

 the reflection, the white is changed to some colour between 

 white and black, t 



To give a correct explanation of passages such as these 

 does not seem to be possible. The splendid colour effects of 

 the Grecian seas may well have excited the imagination and 

 given rise to popular beliefs, with which Aristotle would be 

 acquainted, but the above passages seem to be the results of 

 abstract reasoning. He knew nothing, of course, about the 

 composite nature of white light, and was obliged to make 

 use of some ingenious assumptions to account for colour 

 phenomena. He assumed that minute drops of water, 

 acting as mirrors, may be so small as not to reflect the form 

 of an object, but colour only, such colour depending, in 

 part at least, on the size of the drops. He believed that the 

 " rods " appeared when the very watery clouds, referred to 

 above, varied in density and content of watery vapour, so 

 that the mirrors formed by the minute water drops varied 

 in size. Under these conditions, he considered that, in 

 accordance with the assumptions set out above, coloured 

 streaks of light, e.g., red or yellow, would be produced, for 

 he says : — " The ' rods ' are due to the irregularity of the 

 mirror, not as regards form but colour." § 



Aristotle's explanations of rainbows, primary and second- 

 ary, and their colours are given at great length, chiefly in 

 Meteorol. iii. cc. 4 and 5. Compared with most of his 

 explanations of other natural phenomena, those of rainbows 

 are particularly full, ingenious, and interesting. Read in 

 connection, however, with the more important facts about 



=•= Meteorol. iii. c. 2, s. 6, iii. c. 6, s. 7. f Ibid. iii. c. 6, ss. 1 and 2. 

 I Ibid. iii. c. 4, s. 23. § Ibid. iii. c. 6, ss. 3 and 4. 



