350 



NATURE 



\Aug. 24, i8j?6 



which, in a great manufacturing country like this, ought cer- 

 tainly not to be the least injportant of all. H. T. Wood. 

 Society of Arts, Aug. 22 



The Diurnal Inequalities of the Barometer 



Like the author of the interesting paper on the daily in- 

 equalities of the barometer in Nature, vol. xiv. p. 314, I am 

 one of those who are waiting for the appearance of the second 

 part of Mr. Buchan's essay on this subject. Perhaps the coming 

 meeting of the British Association at Glasgow may elicit from 

 Mr. Buchan the result of his laborious investigations. I own 

 that I am not only anxious to ascertain if his views coincide with 

 my own,i but desire very much to have at my command the 

 thorough discussion of the data for the eighty-six stations which 

 Mr. Buchan has collected. 



So far as a correct explanation of the inequalities is con- 

 cerned, I believe it must be one that can dispense with the 

 lateral movements of the air proposed by Mr. Blanford, and be 

 applicable alike during the calm days of the "doldrums," and 

 during periods of great wind disturbance. It must explain, too, 

 seasonal differences in their amount, and we may infer that what 

 will explain a seasonal difference will probably explain also a 

 geographical difference of the same kind. 



In the barometric co-efficients for Calcutta, supplied by Mr. 

 Blanford, the semicircular one U' is nearly twice as great in 

 April as it is in July, and the quadrantal co-efficient U" is one 

 third greater m March than it is in June. The hour angle u' 

 does not vary so much as it does in this country, and the angle 

 u" shows its usual very remarkable constancy. In England the 

 co-efficient U" seems to have a greater proportionate range than 

 at Calcutta. This will be seen by the following monthly means 

 obtained from Mr. Main's discussion of the observations made at 

 the Radcliffe Observatory, Oxford. 



Mean Daily Quadrantal Oscillation of the Baromettr for each 

 month at Oxford for the sixteen years, 1858- 1873 inclusive. In 

 units of •0001 of an inch : — 



March 



April , 



May 



June 



July 



August 



The epochs of maximum effect seem here to correspond with 

 the greatest thermometric range rather than with epochs of 

 greatest heat. I think it will also be found in this country that 

 this inequality is as large, if not larger, during continuous strong 

 westerly winds as during quiet anticyclonic periods. 



Like Mr. Blanford I was led to this subject by a study of the 

 daily inequalities of the wind. My having arrived at a very 

 different result must be my excuse for pointing out what seem 

 to me to be points of difference between the conditions which he 

 theoretically investigates and those which exist in nature* Mr. 

 Blanford shows that "when a given quantity of heat is em- 

 ployed in heating dry air at the temperature of 80°, it raises its 

 pressure more than seven times as much as when it simply 

 charges it with vapour without altering the temperature." Mr. 

 Blanford very properly premises that this occurs "while the 

 volume remains constant." It is also implied that the; volumes 

 of air are of equal tension throughout. But where do these 

 conditions obtain in volumes of the atmosphere? Such a 

 volume, for example, as rests on a square yard, a square mile, 

 or a hundred square miles of the earth's surface. This volume 

 may easily be supposed to remain perfectly constant, while the 

 tension of its parts may vary enormously. No ordinary addition 

 of heat to the base of this volume will increase its total weight 

 or sensibly add to the tension of the air at the surface of the 

 earth. The added heat vnll alter the relative tension of portions 

 of the lower third or half of the volume, and will be expended 

 in raising to a small extent the centre of gravity of the whole. 

 When this is done, that is, when the dynamical effect of the 

 added heat is completed, the barometer at the base of the volume 

 of the atmosphere will in reality read a little lower, instead of 

 showing the greater tension required by Mr. Blanford's investi- 

 gation. And this will be the case whether the added heat has 

 expanded dry air only, or has evaporated particles of water 

 already in the atmosphere. In either case I apprehend that 

 during the upward movement of the warm air or of the lighter 



' On the Diurnal Inequalities of the Barometer and Thermometer. 

 Quarterly Journal of the Meteorological Society, Oct., 1874 



vapour the barometer would read lower than at the moment 

 when the movement was completed. 



An elevation of the centre of gravity of the atmosphere equal 

 to two-thirds of a mile, barometer at 30 inches, would reduce the 

 weight of the atmosphere by about the one-hundredth of an 

 inch. The centre of gravity of the air over an elevated station 

 like Leh in Ladakh would have to be raised several miles to 

 produce so large a change of pressure as '1034 of an inch, the 

 difference between the maximum night and day value of co- 

 efficient U' as given by Mr. Blanford — so many miles as, in my 

 opinion, to compel one to look for some other cause for the pro- 

 duction of part of the observed effect, and that cause, I believe, 

 will be found in the dynamical one already indicated, 



W. W. RUNDELL 



Visual Phenomena 



Although most people are familiar with the appearances 

 which surround, or perhaps I should say form, the image on the 

 retina of a luminous point, their origin, I believe, is not so 

 generally known, and it is not uncommon to hear there) ascribed 

 to reflection from the eyelids and eyelashes, which in reality 

 plays no part in their production. There are three distinct 

 phenomena which go to make up the appearance of a luminous 

 point, but they are not generally all visible at once. I will 

 describe them for convenit nee of reference as phenomena A, b, 

 and c. 



(a). The luminous point appears to be surrounded by short 

 rays, seldom more than a degree in length, generally much less, 

 the length depending on the brightness of the point and the 

 size of the pupil at the time. 



These rays are what make a bright point look star-shaped 

 (Fig. I). 



(1). Upwards and downwards from the point proceed two 

 bujidles of rays, each often 20° or more in length, and inclined 

 to one another at an obtuse angle (Fig. 2). 



riff. 2 



Fig. I. 



(c). Coloured rays such as are shown in Fig. 3, which are 

 only seen when the eyelids are nearly closed. 



These perhaps it is hardly necessary to say are produced by 

 diffraction through the eyelashes. 



(b) is due to refraction through the small band of tears, which 

 is retained by capillarity in the angle between the inner edge of 

 the eyelid and the eye (shown at t and t', Fig. 4), and which acts as a 

 curved prism, although its effect is only visible when the lids are 

 advanced far enough over the cornea to allow light which passes 

 close to them to enter the pupil. 



The following simple experiments show that this explanation 

 is the right one. 



I. While looking at a bright point so as to see (b), draw down 

 the lower eyelid, the upper bundle of rays will then disappear. 

 This shows that the upper rays are caused by the lower eyelid, 

 and also that as the image on the retina is inverted, the light 

 must take some such course as shown by the dotted lines in 

 Fig. 5. Now in no conceivable way could reflection from the 



