-5- 



NATURE 



[Novi J, 1918 



returning bj the< two gaps at the top and bottom of 

 the vertical lines. One of these gaps is spanned l>\ 

 the simple steel bar to be tested, and the magnetising 

 forces across this sample and across the remaining 

 air-gap are adjusted by means of compensating and 

 tesl coils until they have the same- value. Hence the 

 magnetising force on the bar which is always the 

 difficult thing to measure ran be found, rhe niag- 

 netic induction is measured in the ordinary way by 

 reversing the current in the- magnetising coil and 

 noting the deflection produced on a ballistic galvano- 

 meter in scries with a search coil. It would appear 

 from the tests made by the U.S. Bureau of Standards 

 thai the accuracy of the permeameter is of the order 

 of 5 per cent. Sine the magnetic properties of a 

 strip of transformer steel usual!) vary by this amount 

 at different parts along its length, the accuracy ob- 

 tained is satisfactory. So fai as we know-, however, 

 the Drysdale permeameter is the only one that pro- 

 fesses to test the magnetic properties at different parts 

 of a large block of iron. In this a special tool is used 



to bore into thi iron, so that parts of it can he tested 

 in situ. 



'I'm line of the \ew Zealand Government railways 

 between Christchurch and Timaru is the 1 asiest 

 stretch in New Zealand; it is straight and almost 

 level for nearly the whole distance, but express trains 

 frequentl) involve loads of mure than 400 tons behind 

 the tender, and the prevalent north-west gales make 

 flange resistance heave. The gauge is 3 ft. o in., and 

 the trains were formerh worked by four-cylinder 

 balanced, compound locomotives of the Pacific type. 

 Engineering for November 22 contains an illustrated 

 account of some new non-compound 1 ngines for this 

 service, designed by Mr. II. II. Jackson, chief 

 mechanical engineer. Superheated steam of 1S0 lb. 

 per sq. in. pressure is used, with 54-in. coupled wheels 

 and 17 in. by 20 in. cylinders. The valve gear is of 

 the Walschaert type, with piston valves. The new- 

 engines have been tested against the best compound 

 engines available, and show" a saving of 20 per cent, 

 in wafer and ;; per cent, in coal, including coal used 

 in making up the tin- each morning and during the 

 two hours' stand over at Timaru. Part of the saving 

 is undoubtedly due to the boiler, which is easily the 

 best steaming boiler seen on a locomotive in New- 

 Zealand, and of tliis saving part again is directly due 

 to superheating. The records of the hauling perform- 

 ances are also very good. 



The current issue of Mr. C. Baker's quarterly list 

 of second-hand instruments for sale or hire will repay 

 careful examination by scientific workers. The cata- 

 logue contains descriptions of more than 1500 pieces 

 of scientific apparatus, nearly all of which can be 

 examined at 244 High Holborn, W.C.I. Great pro- 

 minence is given in the fist to microscopes and acces- 

 sdrii astronomical and terrestrial telescopes, theodo- 

 lite-, spectroscopes, projection and photographic ap- 

 paratus, as well as to general physical apparatus. 



OUR ASTRONOMICAL COLUMN. 



Comkts : Wolf's and BoVrelly's. — Ephemeris of 



Wolf's CO 1I1 midnight :■- 



R.A. S. Dect. Log*- Log-1 



Dec. 2 22 im 27 2 53 02003 0-1204 



6 J J 30 22 3 17 0-1996 0-I2Q4 



10 22 41 26 3 36 o-i<)94 0-1367 



14 22 52 34 3 5] 0-1992 0-1480 



t8 23 1 47 4 1 0-1995 0-1576 



22 23 15 246 01672 



26 23 26 II) 4 s 0-2008 01 760 



30 23 37 34 4 5 0-2019 0-1867 

 NO. 256T, VOL. 102] 



I he 1 omel 1 eat bes pei ihelion on Dei ember 13, but, 

 owing to its increasing distance from the earth, ii is 

 not likelj to be brighter than the 12th or 13th magni- 

 tude 



The following is an approximate 

 Borrelly's cornel for Greenwich midnigh 



Li 



Log A 



Dec. 4 7 5 13 31 46 0-141,7 926795 



8 7 3 2 36 20 



1 ■ 7 2 12 ( M 45 0-1545 



"' 7 35 44 59 



20 o 58 5 48 55 o- i6io 9-70 13 



2.1 6 54 1 1 52 25 



28 6 49 50 55 2i, 01687 



The comei is less than half an astronomical unit from 

 the earth during the first half of December. It should 

 be an easy object in small telescopes. The high m 

 declination renders the com, t observable throughout 



the night. 



The Orbit of 83 Aquarii. For the second tim< 

 Dr. R. <1. Aitken has deduced tin orbit of a visual 

 binan star, all the work on which, including the dis- 

 .im(1 measurers, has been carried oul by himself 

 (I. irk Observatory Bulletin No. 317). The star in 

 question is 83 Aquarii (=A4i7), and was noted as a 

 close pair with components of equal brightness i n 

 [902. 1'nti! 1912 the angular motion was nearly uni- 

 form, hut in the last few years tin- motion has been 

 extremely rapid, and the- apparent distance so small 

 as to make the pair difficult to measure- even in. the 

 (6-in. telescope. The provisional elements of the- orbit, 

 with their computed probable- errors, are as follows : — 



P = 23-82 ± 137 years 

 T=i9i7'6S±o 20 year 

 e =0-404 + 0-016 

 a =0245"+ 0-005" 



u> =261-3 ±5'4° 

 i =±56-35 J ± 4 -r 

 0^ = 21-6+2-4- 



The Spectrum of Nova Aquil;e. — An important 

 contribution to the spectroscopic study of Nova Aquilae 

 has been made by- Dr. J. S. Plaskett, who has ob- 

 tained a large number of photographs with a single- 

 prism spectrograph attached to the new 6-ft. reflecting 

 telescope of the Dominion Astrophvsical Observatory 

 at Victoria, B.C. (Journ. R.A.S. Canada, vol. xii., 

 p. 350). Some of the photographs are remarkable 

 for their great extension into the ultra-violet, as mam- 

 as nineteen lines of the hvdrogen series having been 

 measured op June r < > . The description of the changes 

 recorded agrees closely with the accounts given bv 

 other observers. A change towards the nebular stage- 

 was noted on June 20, when the lines 5007, 4685, 

 and 4363 began to show, and on July 11 the nebular 

 stage was strongly marked. The displacement of the 

 first set of hydrogen absorption lines increased at H-y 

 from 205 A on June 10 to 25-4 A on June 24, and 

 from June 10 to June 15 there was a second set 

 having a displacement of 33 A t < > the violet at H7. 

 The measured positions of these- lines agree with the 

 positions computed by Ralmer's formula when the 

 inl is changed from 3646-13 tee 3625-78 for the 

 first and to 3618-4 for the second component on 

 June 15. The position- of -emu sixty-five metallic 

 lines, identified chieffy as enhanced lines of Ti, Fe, 

 Cr, Sr, Sc, and Y. were measured on the- spectra 

 from June to to June 15, and it has been shown that 

 tin- displacements of these were- also proportional to 

 tbe- wave-length, and were exactly the same as those 

 eef lb.- livelmgen absorption lines. Dr. Plaskett con- 

 siders it more likely that tin- displacements an- due t" 

 some phvsical cause acting in the- same manner on the 

 molecules of all the- elements involved than tei veloi 



