456 



NA TURE 



[September 22, 1923 



records the results of inoculation experiments with 

 both Oospora pustulans and 5/>on.'<.v/.,/»,i .uhi, >},tnea, 

 which confirm Owen's original co: Iv'i 



and leave no doubt that the hrst ,, . "U- 



sible for skin spot and the second lor corky bcab. 

 Anatomical investigations of the pustules also show 

 clear differences between those of skin spot and of 

 corky scab, and there is no likelihocxl of a skin spot 

 pustule later masquer.uling as a typical corky scab. 

 Shapovalov's contra iv results were obtained in the 

 United States, and Millard and Burr are therefore 

 led to make the suggestion, inevitably suggested by 

 their own work, that except when the American 

 author examined diseased tubers sent from Europe, 

 he never had typical skin spot under observation. 



Variations in Level of Lake Victoria Nvanza. 



Attention was directed in 1904 to the remarkable 



variations in the level of the Victoria Nyanza by 

 Col. Lyons, who attributed some of them to differential 

 movements in the adjacent land. The general oscilla- 

 tion of the level in that lake and in the Albert Nyanza 

 is described by Mr. C. E. P. Brooks in a Geophysical 

 Memoir, No. 20, issued by the Meteorological Office 

 (1923; 8 pp., I pi.; price, is. 6d.). Mr. Brooks 

 describes the variations in the lake levels as recorded 

 by tide gauges on the Victoria Nyanza from 1896 to 

 1922 and on the Albert Nyanza from 1904 to 1922, 

 and compares the rise and fall of the lakes with the 

 variations in sunspots and rainfall. The discharge 

 from the Victoria Nyanza over the Ripon Falls is 

 estimated at only 6 per cent, of the rainfall on the 

 basin of the lake. Most of the rain is removed from 

 the basin by evaporation, which Mr. Brooks regards 

 as highest during periods of sunspot minima, so that 

 the lake level is then normally lowest. He claims 

 that the lake levels accord more closely with varia- 

 tions of sunspots than with those of rainfall. He 

 points out in illustration of this view that the great 

 rise in the level of the two lakes in 191 7 was "entirely 

 unconnected with any increase in the rainfall." The 

 curves on the plate illustrating the memoir show a 

 general agreement of the sunspot minima with the 

 lake levels : but the agreement is not complete, for 

 the sudden rise in 1901 followed an increase in rain- 

 fall but without any equivalent movement in the 

 sunspot curve. There was a similar disagreement 

 in 1 91 3, and moreover, the high level of the Victoria 

 Nyanza in 1906 preceded instead of followed the 

 sunspot maximum of 1907- 



Space Formula of Benzene, Naphthalene, and 

 Anthracene. — The carbon atoms of the benzene 

 molecule are shown bv B. Orelkin (Jour. Russ. Phys.- 

 Chem. Soc, 1923. 54. PP- 493-532) to be situated at 

 the corners of a regular octahedron. This conclusion 

 is arrived at from geometrical considerations, which 

 show that the above arrangement of the carbon 

 atoms is the only one in which the thirty valency 

 electrons of the benzene molecule can form a stable 

 system. In support of this formula it is claimed 

 that it explains why more or less than six carbon 

 atoms cannot form an aromatic nucleus. The pro- 

 perties of the aromatic nucleus are explained as 

 due to the peculiar arrangement of valency electrons 

 around the carbon atoms, whereby each of the latter 

 possesses two electrons in common with its neighbours. 

 Sachs found that the relative distances of the 0-. 

 m-, and p- positions were as i : s'2 : n'3, and the 

 same proportion is shown to hold for the formula 

 now deduced. The space formulae of naphthalene, 

 anthracene, and chrysene are obtained by the con- 

 densation of two, three, and four benzene nuclei, and 

 the angles of the space lattices of crystals of these 



NO. 2812, VOL. I 12] 



>s OP CoAi ! 1 



iient of Si :• I'M 



! a Ttx.hiii' 



nents in t! 

 '. -..d in v-rt, 



substances are calculated from their n 

 structure. These calculated values agree ver;, 

 with experimental values obtained by other worki 



LOW-TEMPEBATURE CaRH 



Fuel Research Board of tli 

 and Industrial Resean' ■ 

 Paper No. 7 on " Pr^ 



Low-temperature ("arb^ 



Retorts (H.M. Stationery Ofhce, lod. p<> 

 The paper may usefully be read in com 

 of the report of the same body for the yt 

 and 192 1, analysinjj the technical and' • 

 problems to be faced m establishing a Brr trv 



of low-temperatilre carbonisation. The for 



low operating costs, therein emphasised, i^ , 

 minimum of manual labour, and the use of t' 

 tinuous vertical retort is one way of attaining tlii>. 

 An installation of such retorts on the (ilover-Wt-st 

 system now exists at the Fuel Research Station, 

 Greenwich. Though designed for working under the 

 high temperature conditions now current in towns' 

 gas works, they have been employed in carbonisation 

 trials, now reported, in which low working temp>era- 

 tures were maintained. The setting is ill adapted 

 for securing the best results under such conditions, 

 but the tests — admittedly of an exploratory character 

 — have been carried out to obtain information likely 

 to assist in the design of more suitable retorts. 

 Such retorts have been constructed and trials are to 

 be carried out in them. In the present tests flue 

 temperatures ranged from 700° to 850° C, and it was 

 found advantageous to inject steam into the retort, 

 both to cool the coke and to assist in distributing 

 heat through the charge. A coke was obtained 

 containing about 7 per cent, of volatile matter and 

 said to be suitable for use in domestic grates. The 

 high proportion of breeze in the coke suggests trouble 

 and loss in transportation. Per ton of coal, there 

 was obtained a yield of 12-16 gallons of tar having 

 a "low temperature" character and 18-28 lb. of 

 ammonium sulphate. The yield of gas was only 

 45-50 therms per ton — very low from the gas-maker's 

 point of view and fatal to commercial success unless 

 the coke realised a very high price. As no finality is 

 claimed for these results, the results from the new 

 retorts will be awaited with interest. 



Heat Losses through House Walls. — The Build- 

 ing Research Board of the Research Department has 

 issued, as Special Report No. 7, accounts of the tests 

 carried out at the National Physical Laboratory of 

 the heat transmitted through walls of various types 

 when one surface is hotter than the other, of those 

 made in Norway on the heat insulating properties of 

 the walls of experimental huts constructed in more 

 than 20 different ways in use in that countr>-, and of 

 similar tests carried out in Sweden and in Germany. 

 So far as the British tests have been conducted, they 

 show that a solid gravel concrete wall and a wall of 

 sand-lime bricks transmit about the same amount of 

 heat under the same conditions, but that a wall of 

 stock bricks only transmits about J as much heat. 

 A cavity wall of ordinary type transmits about \ to 

 f that of a solid wall according to the size of the 

 cavity. The Norwegian results include the cost of 

 construction and show in a remarkable way the low 

 heat transmission through the less costly wooden 

 walls of various types common in that countr>-. 

 Where cavity walls are used the best arrangement is 

 to place the thicker portion in the interior. The 

 Swedish results, so far as they go, confirm the above 

 conclusions. The German results have led to a sub- 

 division of the air cavity between thin concrete walls 

 into six or more layers by means of paste-boards. 



