272 



NATURE 



[September i, 19 10 



and should be translated to some other sphere." To the 

 man who is interested in every detail, whether scientific 

 or commercial, of the business, each item of knowledge 

 is of positive value, and the more he has the better it is 

 for himself and his employer. 



.Although the use of " invar " wire with a very small 

 coeflicient of expansion in the measurements of the base 

 lines of geodetic survey.s has rendered an accurate know- 

 ledge of the temperature of the wire of less importance 

 than it formerly was, it is still necessary if an accuracy 

 of one part in a million is desired to know that tempera- 

 ture to v/ithin i° C. In general, the temperature of the 

 air, determined by means of a swinging thermometer, has 

 been taken as identical with that of the wire ; but Mr. 

 B. F. E. Keeling, in a communication to the July number 

 of the Cairo Scientific Journal, shows that, under field 

 conditions, the wire temperature is about 2° C. higher 

 than the air temperature owing to the absorption of sun- 

 light af the surface of the wire. His method consists in 

 substituting for the invar wire two wires of copper and 

 constantan of the same diameter as the invar wire soldered 

 together, so as to constitute a thermocouple, and connected 

 to a galvanometer. The temperature of the junction ex- 

 posed under field conditions is then determined from the 

 galvanometer deflections. 



The Builder for August 27 comments on the facilities 

 provided at the Brussels Exhibition for the beginning and 

 rapid spread of fire. The liberal employment of canvas 

 and other textile fabrics as structural and decorative 

 materials doubtless served to render the sections affected 

 by the disastrous fire even more than usually vulnerable, 

 .^part from this characteristic, there is no reason for 

 believing the average exhibition to be much safer from 

 fire than that which has paid so heavy a price for 

 attempted economy. No exhibition within recent times 

 has ueen better conceived from the structural point of 

 view than the Paris Exposition of 1900, where numerous 

 important buildings were constructed entirely in reinforced 

 concrete so as to afford safe accommodation for artistic, 

 historical, and other treasures beyond all price. The 

 example thus set is one that should always be followed, 

 regardless of cost. We must break finally with the tradi- 

 tional flimsy structure of wood, canvas, and plaster. Steel 

 is quite permissible if inexpensively sheathed in fire-pro- 

 tective casing, and is essential in roof construction. But 

 timber wall panels, interior partitions, floors, roof framing, 

 and roof covering must be given up. Expanded metal, 

 wire netting, and steel lathing can easily be stretched 

 between the main stanchions as the basis for incom- 

 bustible walls and partitions formed of cement, mortar, 

 or fire-resisting plaster. 



For nearly four years past, the firm of W'estinghouse, 

 Church, Kerr and Co. has had in hand an investigation 

 of the rust-preventing properties of protective coatings for 

 structural steel. The Engineering Magazine for August 

 contains a brief account of the results in the form of a 

 paper read by Mr. C. M. Chapman before the American 

 Societj' for Testing Materials. More than 500 coatings 

 were tested, each paint being applied to two mild stee! 

 plates of about No. 16 gauge, 2 inches wide and 6 inches 

 long. One plate was given one coat and the other two 

 coats. .After drying, the plates were fastened to boards 

 with galvanised iron tacks, and exposed on the roof at an 

 angle of 45 degrees, facing south. On the day the plates 

 were exposed, scratch marks were made with a sharp 

 instrument across two opposite corners of each plate, 

 NO. 2 13 I, VOL. 84] 



leaving bright metal exposed, so that rusting started 

 irtimediately along these lines. At. regular intervals each 

 plate was examined, and a record made of its condition. 

 The one quality which was being sought was protection 

 against rust. The tabulated results show that with one 

 coating the red leads take first place, both for one year's 

 and two years' exposure. With two coatings and one year's 

 exposure the white leads take best place ; with two years' 

 exposure the red leads are best. Red-lead primers, zinc 

 o.xides, iron o.xides, carbons, and graphites also come out 

 very good in these tests. 



OUR ASTRONOMICAL COLUMN. 

 Astronomical Occurrences in Septbmber : — 

 Sept. I. 8h. 3ni. Venus in conjunction with Moon. (Venus 

 4° 26'?.) 



14. 6h. 501. 10 7h. 23m. Moon occults ai Sagittarii. 



(Mag. 4S.) 

 ,, 7h. 56m. 10 gh. 15m. Moon occults A Sagittarii. 

 (Mng. 4-9.) 



15. Saturn. Major axis of outer ring = 4536", minor 



axis= 13-94". 

 17. I2h. 36nn. Minimum of .\lgol (0 Persei). 



20. 9h. 25m. Minimum of Algol (/3 Persei I. 



21. Ilh. 30m. Saiuin in conjunciion with Moon 



(Saturn 1° 31' S. ) 

 23. loh. 31m. Sun eiitcis Libra. Autumn commences. 

 23. I3h, 43m. 10 I4h. 5m. Moon occulis k' Tauri. 



(Mag. 4-1.) 

 23. I4h. I3ni. to I4h. 59m. Moon occults v' Tauri. 



(Mag. 4-2.) 



The Paris Observ.atorv. — M. Baillaud's report of the 

 Paris Observatory, for igog, opens with a fitting tribute to 

 the memory of M.M. Fraissinet and Paul Gautier, and 

 then proceeds to give an account of each department of 

 the work and its labours during the year, .\mong other 

 things, he mentions that the " cercle m^ridien du Jardin " 

 has been completely overhauled and made perfect, while 

 the programme has been greatly modified ; the instrument 

 is novv' being employed to observe fundamental stars, and 

 will eventually undertake observations for the large cata- 

 logue proposed at the last meeting of the Paris Conference. 



When not in use for spectroscopic work, the large coud6 

 equatorial was employed by MNL Puiseux and Le Morvan 

 for photographing the moon and planets; 21S images of 

 Jupiter tt'ere secured between February and May, and later 

 in the year more than 1200 images of Mars were secured. 



A reproduction of the spectrum of .Arcturus, original size, 

 shows how effective are the new arrangements for employ- 

 ing the large coude for stellar spectroscopy ; the definition 

 is magnificent, and the length of the spectrum between 

 K 4100 and X 4800 is 15-4 cm. 



The small equatorial coude was employed by Dr. Nord- 

 mann in his photometric determination of stellar tempera- 

 tures, and, with JL Salet, he has undertaken the observa- 

 tion of 300 stars of various spectral types. 



Orservations of Comet igioa. — In No. 4433 of the 

 Astronomischc Nachrichten Dr. Karl Bohlin publishes an 

 excellent photograph and two drawings of the great comet, 

 igioa, as observed at Stockholm on January 28. The 

 piiotograph shows the main tail 18° long, cleft at its 

 extremity. The two drawings show enlarged views in the 

 neighbourhood of the head, and in the second the head 

 appears to stand out from the general plane in the form 

 of a half moon. 



The Sun's N'elocitv through Space. — In No. i, vol. 

 xxxii., of the Aslrofihysical Journal Profs. Frost and 

 Kapteyn discuss the value of the sun's velocity through 

 space as derived from the radial velocity of Orion stars. 

 The reasons for employing this class of stars are fully 

 discussed, and the stars considered were taken from within 

 a moderate distance of the apex, or the antapex ; for the 

 former, the position = 269.7°, S= 4-30-8° (1873-0), was 

 taken, and the fact is elucidated that the Orion stars are, 

 as a rule, at a great distance from the sun. This fact may 



