CATALOGUE. — NATURE OF LIGHT. 



319 



the light is atlowcd to pass by both of its parallel edges : but 

 when the light on either side is intercepted, the fringes dis- 

 appear. The crested fringes observed by Grimaldi within 

 the rectangular termination of a shadow, are also shown to 

 depend on the mixture of the two portions of light inflected 

 at the two edges of the object, which form the angle. 



In the second section, the appropriate interval for the 

 brightest light is calculated from experiments of Newton, 

 and from others which are new, and made under a variety 

 of circumstances ; and the measure deduced from each ob- 

 servation agrees with the mean without an error of more 

 than a fourth or a fifth : if the principle had been erroneous, 

 there is no reason why this distance should not have varied 

 at least as much as the measures of the fringes, which were 

 changedinthe ratio of 7 to i, or even in a much greater ratio. 

 There is still, however, some doubt with respect to the cause 

 of the slight difference observed, the measure of the interval 

 being always a little larger in these experiments than in 

 the observations of Newton on thin plates ; and the error is 

 the greater as the tracis of the light is the more rectilinear. 

 The proportions of the intervals for the different colours are 

 also shown to be the same here as in the colours of thin 

 plates : and it is observed that the form of Grimaldi's crested 

 fringes, ought according to the calculation to be that of an 

 equilateral hyperbola. 



The law, being thus established, is in the third place ap- 

 plied to the supernumerary rainbows observed by Dr. Lang- 

 with and others, which Dr. Pemberton has attempted to 

 explain by a comparison with the colours of thin plates. 



The advantage which Dr. Young's explanation possesses 

 is this, that he refers the colours to the light regularly re- 

 flected, and Dr. Pemberton employs the light irregulariy 

 dissipated, of which the effect must be perhaps some hun- 

 dred times weaker. Comparing the two portions of light 

 of which the extreme terminations constitute the common 

 rainbow, he finds that they must cause, by their interfe- 

 rence at other parts, rings of colours, agreeing perfectly with 

 those which were observed in a particular instance by Dr. 

 Langwith, if the drops of rain concerned were all between 

 ■j^ and^Lj of an inch in diameter. 



Hitherto, Dr. Young observes in the fourth section, he 

 has advanced in this Paper no general hypothesis ; and he 

 attempts to infer, by a chain of experimental arguments, 

 that refraction is not produced by an attractive force : since 

 from the smaller length of the appropriate intervals of in- 

 terference in a denser medium, it may be concluded that 

 light moves more slowly as the medium has a greater re- 

 fractive density. He remarks that the existence of the in- 

 tervals of interference in an arithmetical progression, agrees 

 so well with the nature aad properties of a musical sound, 



which consists in the succession of motions in contrary di- 

 rections, at intervals which are also in arithmetical progres- 

 sion, that we can scarcely avoid concluding that the nature 

 of sound and of light must have a very strong resemblance. 



It was conjectured by Newton that the colours of all 

 natural bodies are similar to some of the series of colours 

 produced by thin plates. In this case, as Dr. Young has ob- 

 served in a former paper, they ought to be divided into two, 

 three, ormore portions, by prismatic refraction, as the colours 

 of thin plates necessarily are ; and he has pointe<l out an 

 instance of the kind in the blue light of a candle, which 

 consists, as Dr. Wollaston discovered, of five separate por- 

 tions. He now describes the effect of the prism on the 

 light transmitted by the blue glass sold in the shops, which 

 appears to be divided in a similar manner into seven por- 

 tions. But he confesses that the analogy suspected by 

 Newton is imperfect in more than one respect. 



In the last section an experiment is related by which the 

 effects of thin plates and the general laws of interference 

 are shown to extend to the dark rays discovered by Ritter, 

 and hitherto only known by their effects on metallic oxids. 

 The spectram of rings, which has been repeatedly exhi- 

 bited in the theatre of the Royal Institution, was thrown on 

 a paper dipped in a solution of the nitrate of silver, and the 

 blackening effect was distinctly observable in the portions of 

 three rings, which were marked on the paper, nearly of the 

 same dimensions as the violet rings, but apparently a little 

 smaller. The same mode of analysis, Dr. Young observes, 

 might be extended with great advantage to the rays of invi- 

 sible heat discovered by Dr. Herschel, if we had thermo- 

 meters of suflBcient delicacy to assist us in its application. 



Nature of Light, and Causes of 

 Colours. 



Arjstoteles tie anima. L. xi. c. 7- 

 Zucchii optica philosophia. 1652. 



Maintains the colours are exhibited by transmission only : 

 Kepler showed the same by experiments. Wells. 



Ilooke's miciographla. 



Ilooke's considerations on light. Birch. Ill, 



10. 

 Hooke's opinion of light. Bircii. III. I94. 

 Newton's hypothesis of light. Birch. III. 247, 



278. See Refraction. 

 Mariotte de la nature descouleurs. Oeuvres, 



I. 195. 



