March io, 1898] 



NATURE 



449 



tion which should be of sufticient intensity without a consider- 

 able increase in the size of the illluminating surface. Various 

 attempts have been made to adapt the oxy-hydrogen light for 

 the purpose ; but there always remains the objection, that how- 

 ever small the incandescent portion of the lime may be, it does 



Fig. 1. — Surface markings on Podura scale. Photographed with Swift's 

 1/12-inch apochromatic, projection ocular 2, and central cone. Magnifi- 

 cation, 2500 diameters. 



not emit light of equal intensity over the whole of its surface. 

 This can at once be seen if an image of the lime be projected or. 

 to a screen. The result is uneven illumination, a defect so often 

 seen in high- power photographs, when the image of the radiant 



Fig. 2. — J'lcurosigina angiiiuiuiii. Ph<jtograpbcd u ilh Winkel"* i/20-inch 

 homogeneous immersion, projection ocu'ar 4. Central cone and mal- 

 achite-green screen. Magnification, soto diameters. 



is projected by the achromatic condenser across the object, or 

 what is known as "critical illumination." 



The electric arc is the light which approaches most nearly to 

 an ideal illuminant. The source of lii^ht is extremely small, but 

 the intensity is great, and the incandescent surface is, if working 



under proper conditions, homogeneous. It has until recently 

 been impossible to so control the arc that these conditions 

 could he obtained with certainty. In all forms of lamp, whether 

 hand-fed or automatic, the difficulty has been to maintain a con- 

 .stant position and condition of the crater on the positive carbon. 

 This can be done by having a simple form of hand-feed apparatus 

 with a pin-hole camera attached, through which an image 

 of the carbon points is projected on to a ground-glass screen. 

 Reference lines are provided on this screen, so that the length of 

 arc and position of the positive crater can be continuously 

 observed. The arrangement was exhibited at the two con- 

 versazioni of the Royal Society last year, and has been 

 fully described before the Royal Microscopical Society. With 

 such a form of arc-lamp absolute centiation of the light can 

 be secured and maintained without reference to the micro- 

 scope, after the necessary position of the image of the arc on the 

 screen of the pin-hole camera has been once obtained. The 

 accompanying illustrations have been reproduced from photo- 

 graphs taken with the arc-light so arranged. Fig. i shows the 

 surface markings on a Podura scale, magnified 2500 diameters. 

 Fig. 2 is a frustule of Pleurosigma angu/atiiin, magnified 5cxx) 

 diameters. In neither photograph is there the slightest sign of 

 de-centration, and in both cases centration was maintained 

 entirely without reference to the microscopic image. 



J El»win Barnard. 



Thomas A. B. Carver. 



A METHOD OF MEASURING 

 PRESSURE.^ 



WIND 



'PHERE are few physical problems of greater immediate and 

 obvious practical importance, than that involved in the 

 measurement of air pressures under complex conditions of 

 motion, and there are few problems which present greater 

 difficulty, or — what is worse — uncertainty. It may be com- 

 paratively easy to obtain under any particular set of circum- 

 stances evidences of barometric variation by means of some 

 indicating instrument, apparently suitable for the particular 

 purpose, but it is a very different matter to decide how far the 

 quantitative result is unaffected by actions set up by the instru- 

 ment itself. Thus the record of the pres;!ure plate gives 

 information which is of little, if any, value in relation to the 

 distribution of pressure over a large building ; while the 

 barometer itself is capable of giving misleading indications, 

 whether it is too effectually protected from external influences, 

 or too much exposed. 



For measuring the wind pressure at any point of a structure 

 of considerable size, a receiver or collector is required, with a 

 convenient gauge connected by a tube. It is essential that the 

 collector should not itself give rise to compressions or rarefactions 

 affecting the gauge. To the invention of such an instrument 

 Prof. F. E. Nipher has devoted much attention, and his final 

 apparatus seems to fulfil its purpose admirably. Two equal thin 

 metal discs, 2 '5 inches in diameter, having bevelled rims, are 

 screwed together, so as to leave a small space between, into 

 which a connecting tube is passed through the centre of one of 

 the discs. The end of the tube is flush with the inner surface, 

 and the interspace is filled up with a certain number of layers of 

 fine wire screen, which project at least half an inch beyond the 

 edges of the metal discs. When this simple device is placed in 

 a stream of air, it is found that the effects of rarefaction and 

 coinpression, set up at different parts of the porous screen, com- 

 pletely neutralise each other, so that the pressure at the mouth 

 of the tube is the same as the true intrinsic pressure of the 

 external air. This property of the collector was severely tested 

 by thrusting it out of a carriage window in a train which was 

 travelling at the rate of sixty miles an hour: no effect on the 

 gauge could be noticed, although the instrument was sufficiently 

 sensitive to show instantly the effect of placing the hand at a 

 tangent to the edge. The gauge which Prof Nipher employed 

 was a water manometer consisting of a cylindrical vessel partly 

 filled with water, with a straight glass tube leading out from the 

 bottom and inclined at 5 in 100 to the. horizontal. The open 

 end of this tube was in communication with a collector of the 

 form suggested by Abbe- so. as to .secure a standard pressure of 

 comparison. 



1 "A Method of Measuring ihe Pressure at any Point of a Structure, 

 due to Wind blowing against that Structure." By Franci.s _E. Nipher, 

 {Transactions of the .-^c^emy of Science of St. Loui«, vol. vHi. No. r.): 



- Report of the Chief Signal Officir, 1387, 2, 144. 



NO. 1480, VOL. 57] 



