September 30, 1920] 



NATURE 



'59 



tions the few other known examples described by 

 Matth«?w, Bornemann, and others, and now add* 

 several new genera and species of monactinellid, 

 hexactinellid, octactineiird, and heteractinellid 

 t>pes. Chancelloria, of which the general form '•'*, 

 preserved in spite of the absence of a strong spicular 

 skeleton, upholds Hinde's Heteractinellida as a dis- 

 tinct sub-order. The form o( the spicules, with four 

 to seven horizontal rays and one vertical axial ray, 

 is not tetractinellid, while their grouping is not that 

 of the hexactinellid mesh. The previously known 

 imperfect specimens of this sub-order are of Car- 

 boniferous age. The spicules in th<' Burgess .Shale 

 are usually preserved as pyrite, like those of the 

 British Protospongia, the glories of which are now 

 much diminished by its rivals. 



TKMPERATfREs and humidities in the upper-air con- 

 ditions favourable for thunderstorm development and 

 temperatures over land and sea have been discussed by 

 Capt. C. K. M. Douglas. The results are published 

 as Professional Notes No. 8 by the Meteorological 

 Office. The flight observations were made i < the north 

 of France during 1918 in the preparation of reports 

 and forecasts for military purposes. The most marked 

 feature of the observations was the number of in- 

 stances of an unstable condition above 8000 ft. Every 

 thunderstorm is said to have been associated with a 

 shallow depression or secondary, and many with the 

 arrival of cooler surface-air. Up to a level of 10,000 ft., 

 and possibly above, the higher the humidity the 

 greater the chance of thunderstorms, which are 

 not likely to develop unless there is at least one 

 fairly damp layer as low as 6000 ft. Thunderstorms 

 in our latitudes are not usually severe if the base of 

 the thunder-clouds is above 8000 ft.; the cloud-base is 

 stated to be, as a rule, at about 5000 ft. or 6000 ft. in 

 hot weather, and lower in cool weather. The height 

 of thunder-clouds is said to be very great, the rounded 

 tops attaining 20,000 ft. and the false cirrus 30,000 ft. 

 in hot weather, and 15,000 ft. and 20,000 ft. respec- 

 tively in cool weather. Thunderstorms are divided 

 into three classes : (.A) those due mainly to heated 

 lurface-air in fine, sunny weather; (B) those asso- 

 ciated with powerful upper currents from south-west, 

 the surface-wind being light and variable or south- 

 isterly ; and (C) those associated with very low upper- 

 lir temperatures in the south-westerly or north- 

 westerly currents of cyclonic depressions. Detailed 

 observations and weather-maps are given of each class. 

 I he whole discussion is full of valuable suggestions 

 which will aid alike the weather forecaster and th« 

 .leronaut. 



In the Bulletin de la SocUti d' Encouragement pour 

 I Industrie Nalionale (.May-June, 1920) will be found the 



<:ond part of the experimental study of the quench- 

 ing of carbon steels by MM. Portcvin and Garvin. 

 This study is mainly concerned with the conditions of 

 formation of troostite and martensite. By means of 

 'i>oling curves the authors claim to have cstablish<-d 

 th.it the formation of troostite rorresjxinds to a rapid 

 transformation at about 650° C, whik that of mar- 

 tensite is doe to a relatively slow transformation at a 

 VO. 2657, VOL. 106] 



temperature which is generally below- about 300° C. 

 The former is rendered evident by a clear break in the 

 curve, the latter merely by a gradual change in direc- 

 tion. It was found possible bv choosing a certain 

 rate of cooling to prepare specimens in which both 

 these changes occurred and where the curves showed 

 both tyj>es of change. The authors confirm the 

 generally held view that troostite represents the 

 chemically stable mixture of iron carbide and a-iron 

 in a fine state of division, but their results do not 

 throw any fresh light on the constitution of martensite. 



A VERY convenient new showroom is being opened 

 by the Cambridge and Paul Instrument Co., Ltd., at 

 the commodious head office into which the firm has 

 just moved at 45 Grosvenor Place, S.W. i. .As a 

 nucleus of the fine assemblage of scientific apparatus 

 which is being brought together there, the collection 

 of instruments shown by the company at the Engineer- 

 ing Exhibition at Olympia has been provided with a 

 permanent home. This relates more particularly to 

 industrial apparatus, and one of the most interesting 

 items is the firm's new CO, recorder, which is entirely 

 electrical in its principle. It depends for its action on 

 resistance measurements revealing the difference in 

 the rates of cooling of two platinum helices, one 

 exposed to air saturated with water-vapour, and the 

 other to the llue-gases to be tested. The indicating 

 instrument is arranged so as to read direct in per- 

 centage of COj, and can be at a considerable distance 

 from the gas-chambers. Continuously recording 

 apparatus can be used, or one indicator can be 

 arranged to be plugged on to any one of a number of 

 CO, meters. \ variety of patterns of both resistance 

 and thermo-electric thermometers and pyrometers are 

 included in the collection, as well as temperature 

 measuring apparatus on other principles, pressure- 

 gauges, measuring microscopes, fuel calorimeters, and 

 electrical testing apparatus. 



An interesting report of tests on a 1500-kiIowatt 

 Ljungstrom' steam turbin<' conducted by Capt. H. 

 Riall .Sankey appears in Engineering for September 10. 

 The turbine had been in regular service for fifteen 

 months at the date of the test in the works of Messrs. 

 .Marshall, Sons, and Co., Ltd., Gainsborough, and 

 was constructed by the Brush Electrical Engineering 

 Co., Ltd., at Loughborough. No overhaul of the 

 turbine or any special preparation was made for the 

 trials. The principal result, corrected to the guarantee 

 figures of 190 lb. per sq. in. stop-valve pressure, 

 634° F. stop-valve temperature, and 2875 in. vacuum 

 at the turbine exhaust, was 11-95 'o '^■"" ">- of 

 steam per kilowatt-hour, as measured at the switch- 

 board. The result is slightly below the original 

 guarantee, and indicates that there can h.-ive been no 

 serious wear of the blading during the period of ser- 

 vice of the turbine. The heat supplied per minute 

 per kilowatt ranged from 2580 to 2570 thermal units, 

 and the overall efficiency ratio varied from 680 to 

 684 [K-r cent. 



The steaming trials of the armoured cruiser 

 1 1. M.S. Raleigh have just been completed on the 

 Clyde, and form the subject of an article in 



