430 



KNOWLEDGE. 



November, 1910. 



CLOTH FROM BAN.\N.\ FIBRE.— .\ specimen of Cloth 

 made from banana fibre, recently exhibited at the fair at 

 Chunking, has been sent by the British Consnl of that place 

 to the Board of Trade. 



It is prepared in China by unrollinsi the stall<s of the 

 banana when about a year old. and steaming them over boilint; 

 water until qnite soft. The green outer skin may now easily 

 be stripped off by passing the stalks through a machine 

 provided with two blunt scrapers, the residue then containing 

 the fibres. The material is ne.xt wrapped in a cloth and 

 hammered so as to separate the fibres and expel the moisture, 

 and the fibres are then pulled apart and spun into threads for 

 "Caving. The method of separation is thus very similar to 

 that used in the preparation of ramie fibres. 



So far only a few lengths of cloth have been woven, and the 

 price is therefore high (/.I 3s. fid. for a roll 15 feet long by 

 3 feet broad), but with increased demand the price will fall 

 and the material will probably be extensively used in the 

 manufacture of clothes for summer wear. 



The ultimate fibres, the microscopical form of which is here 

 shown, are about fifteen inches 

 in length and \ery strong. 

 It will be noticed that the 

 twisted form of the bottom 

 fibre in the photo-micrograph 



recalls the characteristic twist ^^^^^^^^^^■■^■^L 

 of the cotton fibre, from 

 which, however, it may be 

 distinguished by not having 

 the twists along its whole 

 length. The jointed structure 

 is similar to that of the linen 

 fibre, which it also resembles 

 in its general chemical 

 characteristics and behaviour 

 towards difi"erent reagents. 



A specimen of the pali 

 yellow cloth, for which the 

 writer is indebted to the Com- 

 mercial Intelligence Branch 

 of the Board of Trade, was 



exceedingly tough and apparently very durable. It contained 

 10'5 per cent, of moisture, and yielded, on ignition, F? per 

 cent, of mineral matter (ash), the latter figure being very 

 similar to that given by linen, which usualK- contains about 

 F5 per cent, of mineral matter. 



CHEMISTRY AT THE BRITISH ASSOCIATION.— 

 The reports of various Committees to Section B (Chemistry) 

 at the recent meeting of the British .Association at Sheffield 

 included those on the " Study of Hydro-aromatic Substances." 

 on " Electro-.Analysis." on " Dynamic Isomerism." and on 

 ■'The Transformation of Nitro-amines. etc." all of which 

 are interesting only to the chemist. 



A report on " Gaseous Combustion," by Dr. Rone. gi\ es an 

 extremely interesting summary of the chemical researches upon 

 the nature of gaseous combustion during the last 30 years. 



Of the papers read before the Section mention may be made 

 of that on "The Influence of Chemical Composition and 

 Thermal Treatment on the Properties of Steel," by Professor 

 McWilliam, and that on "The Crystalline Structure of Iron 

 at High Temperatures," by Dr. Rosenhain, who shows that 

 between the ordinary temperature and 1000° C, iron exists in 

 three distinct modifications, possessing very different physical 

 properties. There was also a contribution by Dr. Copeman 

 on " Ferro-silicon," the dangers attending the transport of 

 which have already been described in these columns. 



The paper by Dr. Howe on " .-Mlotropy or Transmutation " 

 is a plea for regarding the (juestion of transmutation of the 

 elements from a different point of view than the usual one. 

 The allotropism of the two forms of carbon, diamond and 

 lampblack, which may be transmuted into each other, is 

 conunonly explained on the ground that they are the same 

 element because the> \irld the same products of decomposition. 



FiGL KL 



In place of the statement that "the elements cannot be 

 transnuited into each other." it is suggested that it would be 

 more philosophical to say. " Hitherto no elements have been 

 transmuted into each other except those which transmute so 

 readily that the derivatives of only one of them ha\e been 

 recognised." Such a point of view would render the transmu- 

 tation of an element, say copper, into another, say, lithium, 

 not so inherently improbable as to demand exceptionally 

 conclusive evidence. 



FXONOMIC BIOLOGY. 



By Walter E. Coi.linge, M.Sc, F.L.S., F.E.S. 



THE CLASSIFICATION OF BACTERIA.— In the 

 investigation of plant diseases due to bacteria, great difficulty 

 is frequently presented to the investigator in identifying the 

 species under observation. This diificulty of arriving at a 

 conception of species has been experienced in other branches 

 of biological science, and has led to the recognition of groups 



of related species, such groups 

 being treated as units until 

 such time as the progress of 

 knowledge would allow them 

 to be broken up into their 

 component species. 



The early work with bac- 

 teria was \ery largely con- 

 fined to pathogenic forms, 

 and pathogenicity was relied 

 upon to define the limits of 

 the species, but as the study 

 extended to non-pathogenic 

 forms, reliance was placed on 

 various other physiological 

 reactions, singly or in com- 

 bination, although, unfortu- 

 nately, there has been little 

 agreement amongst workers 

 as to the relative value of the 

 different reactions recorded. 

 Recognizing the importance of a uniform and concise 

 metliod of recording such reactions the Society of American 

 Bacteriologists adopted an official classification card for this 

 purpose. An important part of this card was the " group 

 number." in which the results of ten different reactions were 

 expressed numerically. 



CONSTANCY OF CHARACTERS.— As the usefulness 

 of the group number becomes recognized there is a growing 

 desire to extend its range until it shall classify cultures as 

 closely in accord with the idea of species as possible. It 

 therefore becomes imperatively necessary to know both the 

 constancy of these reactions and the extent to which the 

 group number can be followed in classification without 

 separating various strains of the same species. 



In this connection an excellent piece of work has recently 

 been published by Mr. H. A. Harding (Tech. Bull. No. 

 13, N.Y. .\gric. Exp. Station), who has selected an organism 

 {Psciidoiuonas caiiipcstris) the limits of which are clearly 

 defined ; he has also studied a large number of strains under as 

 wide a variation in conditions as could be reasonably expected 

 to occur in ordinary laboratory work. The study has 

 extended o\-er a year and a half, using media prepared by 

 difterent workers, and in some cases the observations were 

 made by three different workers separately. In short, the 

 effort was made to find the maximum variation which could 

 be expected where obser\ations were made in accordance with 

 the official directions, or with the deviation therefrom which 

 could be reasonably expected in practice. 



CONCLUSIONS. — The results obtained are set forth in 

 great detail, and show that the card system offers a basis for 

 classification which for convenience of application and in 



111. 



