i66 



NA TURE 



d 



[June 17, 1S97 



If = 1000, the loss is 54 per cent. The me thod employed in the 

 investigation is that suggested by Lamb and Niven ; the author 

 adds an expression for n, the solid angle subtended by the 

 circuit at any point, in terms of Bessel's functions. Mr. Ever- 

 shed referred to some experiments of his own, from which he 

 concluded that the author's formula gave too low an estimate of 

 the attenuation ; the discrepancy indicated that some term had 

 been neglected. Mr, Yule doubted whether the equations given 

 by the author were quite applicable to sea-water. There was 

 need, apparently, of a term involving the polarisation of the 

 medium. Mr. Heaviside communicated a criticism of the 

 paper. It was not necessary to investigate the problem for any 

 particular form of circuit from which the waves proceed. The 

 attenuating factor for plane waves, due to Maxwell, was suffi- 

 cient. Taking the l^est-known value for the conductivity of 

 sea-water, there was no reason why the conductivity should 

 interfere with signalling. A considerably greater conductivity 

 must be proved for sea-water before it could be accepted that 

 the failure of experiments on telegraphic communication with 

 light-ships from the sea-bottom was due to that factor. It was 

 unlikely theoretically ; and Mr. Stevenson had contradicted it 

 from a practical standpoint. For some reason, the account of 

 the light-ship experiments had not been published, so that there 

 was no means of finding the real cause of failure. — Mr. T. H. 

 Blakesley read a paper on a new definition of focal length, and 

 an instrument for determining it. The author asserts the 

 principle that the focal length of a lens-combination is an abstract 

 quantity, not necessarily the distance between two particular 

 points. It is a quantity best defined in terms of some function 

 of the two distances of object and image from their appropriate 

 focal centres. Such a function is the magnification factor, m, 

 the linear ratio of image to object, positive if the image is erect 

 with regard to the object. Consider a particular pair of conjugate 

 foci on the axis of a lens-system. Let one of these foci be at 

 distance v from some fixed point on the axis, measured posi- 

 tively, in the direction of the rays. Then — is constant, and is 



am 

 the focal length,/ Ifv^ is the value of z/ when ;« = o, v — V(,=/.m. 

 Let t( be the position of the other focus, and «„ its value when 



m = 00.' Then u — w^ =/.7/i-^ ; and ~Z^ = m'^. The last 



expression, /n^, may be called the " areal magnification " ; it is 

 important in determining photographic exposure. The author 



describes an optical bank which enables ^ to be measured by 



dm 

 a very simple operation ; it gives also a record, on a paper strip, 

 of the magnification-factor corresponding to various relative 

 positions of object and image. Dr. S. P. Thompson said the 

 paper was the most important contribution to geometrical optics 

 that had appeared for many years. The introduction of the 

 magnification function was a most useful device leading to ex- 

 ceedingly simple results. The important thing to measure was 

 not so much the focal length as the reciprocal of that quantity. 

 Dr. Chree said the photographic method at present used at 

 Kew for determinations of focal length gave greater security 

 than any more direct method. The colour of the light had to 

 be taken into account. Mr. Blakesley, in replying, called 

 attention to the use of his strip diagrams of magnification, for 

 enlarging purposes in photography. When the magnification 

 along some definite line vvas Vnnwn, the focussing-clolh might 

 almost be dispensed^with^Dr. J. A. Fleming read a paper on 

 a method of determining magnetic hysteresis loss in straight iron 

 strips. The author's process is based upon the use of the bifilar 

 reflecting watt-meter. The samples of iron, large or small, in 

 the form of straight strips are inserted in a long solenoid. The 

 solenoid is traversed by an alternating current, and the square- 

 roots of the mean-square values of the current are determined by 

 a Kelvin balance. A flat bobbin of fine wire may be slid along 

 the strip ; an electrostatic voltmeter connected to the ends of this 

 exploring coil gives the square-roots of the mean-square values 

 of the electromotive force in that coil. From these measurements 

 and the knowndimensionsof the solenoid and coil, the induction 

 density, B, can be found at any point of the length of the strip. 

 From these results a curve is drawn, coordinating the values of 

 B to corresponding distances along the half-length of the strip. 

 Assuming the hysteresis loss per cycle, per c.c. of iron, to vary 

 as the I "Sth power of the maximum induction density, and then 

 raising all the B ordinates to the i -eth power, and plotting a 

 new curve over the first, another curve is obtained which repre- 

 sents the variation of hysteresis loss per c.c. of iron from point 



NO. 1442, VOL. 56] 



to point along the half-length of strip. Now, at some point 

 along the half-length of strip there must be a section where the 

 induction density is B, such that the true mean hysteresis loss for 

 the 7vAo/e bar is proportional to B/'^. Let this value of the 

 induction density be called the "effective value" and the corre- 

 sponding point in the strip the " effective point." Let M-Bi« 

 stand for the mean ordinate of the curve representing the varying 

 values of Bi'« all along the half-length. Then, evidently, 



B, 



VMB' 



The following curious experimental result is found. What- 

 ever may be the length or section of the iron strip, the point at 

 which the actual induction density has a value equal to the 

 "effective" value, always comes at the same proportional dis- 

 tance from the centre of the strip. This distance is very exactly 

 equal to 0-56 of the half-length, as measured from the middle ; 

 or 0-22 of the whole length from one end. If, therefore, the 

 secondary coil is placed at that spot, and the secondary voltage 

 then observed is used to calculate the induction density, the 

 value so obtained corresponds to the true mean value of the 

 varying hysteresis loss per c.c. all along the strip. Mr. Carter 

 asked whether roots other than the i •6th gave a similar constant 

 value of the induction density. Dr. Fleming said it seemed to 

 be the result o£_ac ^dcnt th^t the i 6th root gave a constant 

 valueJorjran-^The President proposed a vote of thanks to the 

 autKors,"and the meeting adjourned until June 25. 



Geological Society, May 26.— Dr. Henry Hicks, F.R.S., 

 President, in the chair.— On augite-diorites with micropegmatite 

 in Southern India, by Thomas H. Holland, Officiating Super- 

 intendent, Geological Survey of India. This paper dealt with a 

 series of basic dykes intersecting the pyroxene-granulites and 

 gneisses of the Madras Presidency, and believed to be of the 

 same age as the lava-flows of the Cuddipah system. These 

 dykes consist essentially of augite (near hedenbergite) and a 

 plagioclase-felspar (near labradorite), between which we find 

 masses of micropegmatitic intergrowths of felspar and quartz, 

 with a micro-miarolitic structure. Around the patches of micro- 

 pegmatite, chemical changes have frequently taken place in the 

 minerals of the rock. After discussing the chemical constitution 

 of the rock, and of its various constituents, and the relation 

 between the micropegmatite and the surrounding minerals, the 

 author pointed out that three methods for the formation of the 

 micropegmatite may be conceived of: (o) during the primary 

 consolidation of the magma ; (;8) by secondary changes induced 

 in the rock ; (7) by subsequent intrusion of granophyric material 

 into the augite-plagioclase rock. In opposition to (7), the 

 author pointed out the entire absence of granitic intrusions in 

 the neighbourhood. He regarded the absence of all proofs of 

 subaerial hydration, and the remarkable freshness of the rocks as 

 precluding the possibihty of the micropegmatite having been 

 formed by secondary change. The primary origin of. the micro- 

 pegmatite he believed to be proved by (i) the crystallographic 

 continuity of its felspar with that of the normal plagioclase of 

 the rock ; (2) the mode of occurrence of the micropegmatite, 

 filling in the angles and spaces between the augite and the 

 plagioclase ; and (3) its variation in coarseness of grain agreeing 

 with that of the remaining two constituents of the rock. An 

 interesting discussion took place upon the question of the 

 primary or secondary origin of the micropegmatite in basic rocks. 

 — The laccolites of Cutch and their relations to the other igneous 

 masses of the district, by the Rev. J. F. Blake. The author has 

 observed thirty-two domes of various kinds in Cutch, distributed 

 as follows : (i. ) those connected with the northern islands; (ii.) 

 those of Wagir ; and (iii.) those along the northern edge of the 

 mainland. They are divisible into four classes : {a) those which 

 are so elongated on the line joining adjacent ones that they seem 

 to be mere modifications of anticlinals, though the supposed 

 anticline is not really continuous ; {b) those which lie in a line, 

 but are not elongated in that direction, and often in no other ; 

 {c) those which are related to a fault, which cuts them in half; 

 and [d) those which are not in any particular relation to each 

 other, or to any other stratigraphical feature. The domes varied 

 in degree of perfection : some were irregular, while some had 

 the strata running in concentric circles, the outer and newer 

 strata dipping away from the inner and older. In no less than 

 ten of the thirty-two domes igneous bosses were found occupy- 

 ing the centre, and these were distributed amongst all of the 

 above classes. The author gave reasons for maintaining that 

 the domes were the results of intrusion of igneous rocks in the 



