240 BELL SYSTEM TECHNICAL JOURNAL 



The second class of fibers, coming from the vegetable world, are 

 alike in being composed of cellulose, a substance like the protein in 

 having a high molecular weight but unlike it in that its polar groups 

 are hydroxyls which have a faintly acidic rather than amphoteric 

 nature. These are the groups in cellulose with which water is likely 

 to associate itself. Such data as are available concerning vegetable 

 fibers other than cotton, notably linen, ramie, manila hemp, and wood 

 pulp, indicate a strong resemblance not only chemically but hygro- 

 scopically and electrically. 



The subclass embracing only cellulose acetate as a commercial 

 fiber is chemically more neutral and non-polar in type than other 

 cellulose fibers, with which fact it is reasonable to associate its lower 

 hygroscopicity and consequent better electrical characteristics. It is 

 probable that cellulose nitrate and cellulose ethers will be found to 

 fall in this class but artificial silks other than cellulose acetate absorb 

 more water and appear on chemical grounds to be better classified 

 with the cellulose fibers of natural vegetable origin. 



It cannot be decided from available information whether the 

 similarities and differences in electrical properties among the textiles 

 are traceable directly to chemical similarities and differences or 

 indirectly to physical (colloidal) structures which in turn are deter- 

 mined by chemical composition. In either case it is possible to 

 account for the high sensitivity of all the fibers to moisture and the 

 variation in sensitivity from species to species by assumptions as to 

 the distribution of water in them. Water which collects in any 

 isolated form in the material will have little electrical effect compared 

 with that which forms continuous filaments. The distinction we are 

 making is essentially the same as that of Evershed when he referred 

 to part of the water as "dormant," though we do not attach the same 

 importance as he does to electrokinetic redistribution of water under 

 electric stress. Each increment of water may be considered as under- 

 going partition into two portions, one causing a large increase of 

 conductivity and one having a negligible effect, in a ratio determined 

 in some fashion by the structure or nature of the material and the 

 humidity of the atmosphere. The ratio of the two portions which 

 are in equilibrium via the surrounding atmosphere will be subject to 

 constant readjustment under changing conditions. 



The fact that the electrical characteristics of the two classes of 

 fibers as affected by moisture appear to be specific properties of the 

 substances involved suggests some highly regular distribution pattern 

 of conducting water paths determined by the chemical or physical 

 (colloidal) structure of the material. Such a regular pattern may 



