June 9, 192 1] 



NATURE 



457 



with the surface at P. Since this ang^le is directly 

 projx»rtionaI to the distance between the two wave 

 surfaces, the normal at any point — that is, the direction 

 of the ray — varies at a constant rate, and is therefore 

 the arc of a circle. If vj. — x and R^ is the radius 

 of this circle (which may be called the refractive 

 radius), 



Rr^— =.i-, so tliat Rr= . 

 H a 



In geographical miles H = 4-32 about, which makes 



Rr= 14,900 miles. 



The course of the ray is the same as it would be 

 if it passed through an infinite number of vertically 

 placed acute prisms of height H, having^ a refractive 

 index fiaj(i — a), with their bases occupying the whole 

 surface of the ground. Since for horizontal rays 

 these prisms are in the position of minimum deviation, 

 rays which are pointed a few degrees up or down 

 will still be arcs of the circle with 14.900 miles radius. 

 The usual tables for the distance of the sea horizon 

 assume that the horizon is .k: miles distant, when the 



height of the eye (//) = '— ~, R^ being the earth's 

 2 Rf 



radius. 



If refraction is taken into account, 



2^R, R/' 



or, in numbers, without refraction, 



with refraction allowed for, 



A=o-67xK 



Thus the sea horizon viewed from a height h, or a 

 mountain of this height just visible from sea-level, 

 is a good deal further off than the ordinary tables 

 would indicate. 



If an atmosphere of the same height and density 

 as that of the earth covered a globe of 14,900 miles 

 radius, an elevation at any one point of its surface 

 would be visible from every other point, and a light 

 at one end of a diameter would appear to an observer 

 at the other end as a bright line extending round the 

 whole of his horizon. A. Mallock. 



9 Baring Crescent, Exeter, May 12. 



Young's Interference Experiment and the 

 Spectrometer. 



In Nature of April 28, p. 268, Dr. R. A. Houstoun 

 directs attention to the use of the spectrometer for 

 Young's double-slit experiment. In a letter on "The 

 Visibility of Interference Fringes and the Double 

 Slit " (N.4TURE, July 26, 1917, vol. xcix., p. 424) the 

 present writer made reference to a similar optical 

 arrangement. In that, letter emphasis was laid, not 

 on the advantages of the method for observing fringes 

 and evaluating wave-length, but on its use for study- 

 ing the changes in the visibility of fringes which 

 occur as the width of the spectrometer slit is altered. 

 In view of Prof. Michelson's recent use of the double 

 slit for the measurement of the angular width of dis- 

 tant stars, I may be pardoned for directing attention 

 to my note of some vears ago, and for pointing out 

 the ease with which an experiment similar in method 

 to that of Prof. Michelson may be performed by means 

 of an ordinary spectrometer. It is true that, instead 

 of using a source of fixed (but finite) width and a 

 variable double slit, the converse arrangement was 

 emploA'ed, but in principle the methods are identical. 

 It might be worth while, however, to van,' the experi- 

 ment by replacing the spectrometer slit by a small 

 circular aperture and using a double slit of variable 

 width. 



NO. 2693, VOL. 107] 



May I also point out that the spectrometer may be 

 used advantageously for an experiment on the limit 

 of resolution of a telescope? It will be recalled that 

 in the standard experiment a distant piece of gauze 

 of fairly large mesh is viewed through a telescope 

 before the objective of which an aperture is gradually 

 narrowed until one set of wires disappears. If the slit 

 of the collimator of a spectrometer be removed and 

 in its place a piece of gauze of fine mesh be sub- 

 stituted, a much more convenient arrangement is 

 available. During the past winter I found that 

 students make quantitative measurements with such 

 an arrangement without even darkening the labora- 

 tory. 



In conclusion, it is well to note that for much of 

 the above work even a spectrometer is not necessarv. 

 A telescope, a good lens, and adjustable slits are the 

 onlv essentials. John K. Robertson. 



Queen's University, Kingston, Canada, 

 May 12. 



The Reparation Act and Scientific Research. 



The columns of Nature could, no doubt, be crowded 

 with complaints concerning the Reparation Act, but 

 as a sufferer I may be permitted to give the following 

 instance of the extraordinary way in which this Act 

 is used to hinder research. 



I ordered a case of chemicals from Germany for 

 myself and other workers in the Biochemical Labora- 

 tory, Cambridge, on February 2. In order that these 

 should not come under the Reparation Act, they were 

 dispatched on March 24 and arrived on April 8. 

 Although the Reparation Act did not come into force 

 until April 15, the goods were seized at Grimsby, and 

 after a week's delav I was asked for all the originar 

 documents showing that the goods were ordered before 

 March 8 and delivered before April 15. These were 

 at once sent to London, but no reply was received 

 from the Customs until I was forced to request the 

 Medical Research Council, for whose work the 

 chemicals were required, to apply to the Customs to 

 free the goods as soon as possible. After three weeks' 

 delay I received a letter from the Customs saying that 

 the original documents were insufficient, and that a 

 statutory declaration was required to confirm the par- 

 ticulars and to prove that the contract had not been 

 fulfilled. This necessitated two visits to a commis- 

 sioner of oaths and the preparation of a lengthy manu- 

 script document adorned with red seals, the cost of 

 which I have still to discover. This evidence was 

 forwarded to the Customs a fortnight ago, and I have 

 received no answer. The goods are still at Grimsby, 

 the work of sev-eral people is being delayed, and the 

 goods will apparentlv remain impounded (although 

 thev never came under the Act) unless I am prepared 

 to submit to what might almost be called blackmail. 

 To obtain the chemicals I must pav the full 50 per 

 cent, myself, the funds of the Medical Research 

 Council being unavailable for the purpose, and I 

 must trust to recover the monev from the Customs 

 when my claim has been recognised — evidently a verv 

 doubtful eventualitv. 



All this delay, the expense, including the commis- 

 sioners of oaths, not to speak of the worrv and waste 

 of time of an interminable correspondence, are ap- 

 parentlv due to nothing but the red tape of a Govern- 

 ment oflfice. H. Onslow. 

 May 27. 



British Laboratory Ware and Chemicals. 



The question of the quality, supply, and prices of 

 British laboratory glassware, porcelain, and chemicals, 

 including research chemicals, is under consideration 

 bv a committee of the British Science Guild. The 



