I068 G. ROSSI 



this specific reagent. It may be contended that in view of the affinity 

 between pectic substances and carbohydrates any reagent attacking the 

 former strongly must necessarily attack the latter, at any rate slightly. 

 If in order to surmount this difficulty an insufficient quantity of reagent 

 were employed, the resulting separation of the fibres niight be insufficient. 

 That is why all chemical means tend to break up the fibre. A similar 

 defect is obviously inherent in ordinarj^ microbiological methods, i. e. those 

 practised in rural districts for retting purposes. 



In these latter it is not only the bacteria of pectic fermentation which 

 multiply, but also other bacteria, chiefly those of cellulose fermentation. 

 This happens so easily that there is the risk of the fibres being injured in 

 the same way as by the action of a chemical reagent common to the two 

 substances. 



If it so happens that specific pectic bacterial ferments develop ex- 

 clusively, or predominate over the others for some length of time, it may 

 be assumed that microbiological retting will take place without any draw- 

 backs ; but in this case only. Moreover, it is still necessary to stop the retting 

 process by taking the textile materials out of the liquids in which they 

 are steeped when the necessar}^ limit has been readied. This is not always 

 easy, as it is by no means rare for a lot of hemp or flax to be classified 

 as " over-retted ", depreciating markedly in consequence; the most able 

 peasants may make mistakes in this respect. This being the case, nothing 

 but a ferment, fixed or soluble, specific to the pectic substances, acting in 

 pure culture if it is an organism and alone if it is an enzyme, wordd over- 

 come all the difficulties, because the degree of sensitiveness required for 

 this very special problem can only be obtained from a biological action. 



With the help of other workers I showed that there are micro-organisms 

 which attack plant tissues but leave the cells and their derivatives intact. 

 Thus, on preparing a culture of Bacillus Comesii in the presence of a leaf 

 of Medicago lupiilina or Coronilla Emerus (sterilised by means of special 

 methods leaving the leaf chemically intact), these leaves are decomposed 

 into 3 parts : 



(a) the parenchyma cells which fall to the bottom of the medium ; 



[b) the sclerenchyma, which remains intact in skeleton form ; 

 {c) the epidermis, which floats on the liquid. 



This decomposition goes no farther. By preparing a culture of the 

 same bacillus in the presence of wisps of hemp, under the same conditions 

 as the leaves, the dissociation of the stalk into its three parts is secured : 

 wood, textile bast-fibres and cells of the cortical layers. All these parts, 

 however, and what is more important, the fibres, may be kept in the pre- 

 sence of the still living bacteria for as long as 2 years without being at- 

 tacked and even without the breaking-up of the primary fibre. The cellulose 

 wall thus cannot serve as food material for these bacteria. 



All these facts (utilised in the method of retting bearing my name) con- 

 stitute the pectic fennentation referred to above. 



The proposed microbiological methods of retting are many, but can be 

 brought under 2 heads : i) those in which the addition of a selected fer- 



