548 Transactions of the Society. 



A dendrite appears to continue its growth as long as the 

 nucleus remains in part unoxidised. A dendrite in a cotton 

 paper, from a book dated 1850, showed neither increase in size nor 

 change in contour during a period of five recent years, for the 

 simple reason that the nucleus had become completely oxidised, 

 probably many years ago. It would appear from this, that after 

 the central nucleus of a dendrite is completely oxidised, the oxides 

 produced do not of themselves spread to any appreciable extent ob 

 the fibres. The growth of a dendrite is more active and rapid in 

 its initial stages, before the nucleus becomes thickly coated with 

 oxide. The dendritic growth creeps along the various fibres in 

 characteristic fashion. The more or less cylindrical fibres of straw, 

 esparto, and linen, become sheathed in copper oxide, the deposit 

 upon linen being usually thick and rather patchy. The wider, 

 tubular fibres of cotton cellulose, become coated in a similar 

 manner to those of linen, but when the cotton fibres are more flat 

 and riband-like, the dendrite often spreads out laterally from fibre 

 to fibre in beautiful fern-like traceries. The latter habit is typical 

 also of almost all dendrites found in papers composed of wood 

 cellulose, the flattened fibres of which seem to present an ideal 

 surface for the dendritic growth, the oxide of copper spreading and 

 branching from fibre to fibre in a beautiful and delicate crystalline 

 pattern. As pointed out by Scales, the ramifications of dendrites 

 generally depend upon the lie of the fibres in the paper. This is 

 true for straw, esparto, linen, and some cotton papers ; but in the 

 case of wood cellulose papers, and some papers composed of flat- 

 tened cotton fibres, the crystalline energy of the growth asserts 

 itself, allowing of frequent lateral growth in a direction often at 

 right angles to the lie of the fibres. This fact appears to be due 

 partly to the flat shape of the fibres, and partly to their compara- 

 tive softness, which allows of a more vigorous growth. 



Most dendrites, especially old ones, are composed almost 

 wholly of black cupric oxide, but in many cases, especially in 

 recent dendrites, I have observed the formation of dark-red, trans- 

 lucent cuprous oxide, and of a yellowish substance similar in 

 appearance to a partially hydrated cuprous oxide. From a careful 

 consideration of many dendrites during various stages in their 

 growth, I have come to the conclusion that the copper is conveyed 

 along the pores of the cellulose in a hydrated cuprous condition, 

 possibly in chemical union with the cellulose, and that it crys- 

 tallises in the form of capillary-aggregates of cuprous oxide upon 

 the surface of the fibres, the red cuprous oxide being subsequently 

 oxidised to the black cupric form. I have not at any time 

 observed the characteristic green colour of cupric carbonate in 

 dendrites, but, nevertheless, it is most probable that the minute 

 quantities of carbonic acid, and perhaps, of ammonia, present in 

 the atmosphere, play an important part in the oxidation of the 



