THE FERMENTATION OF CELLULOSE. 195 



POPOFF (I.) in 1875, and more accurate researches on the same 

 point were published by HOPPE-SEYLER (I.) in 1886. The latter 

 kept, with exclusion of air, either mud from marshes and rivers, 

 or else clean paper inoculated with a little mud and then distributed 

 in water, and demonstrated that in the mixture of gases evolved 

 therefrom methane predominated greatly at the outset, but there- 

 after gradually diminished to the proportion i : i, so that Hoppe- 

 Seyler, being unable to discover any other decomposition products, 

 expressed the opinion that the cellulose is hydrolised by the action 

 of the bacteria which from microscopic examination he asserted to 

 be the same as Yon Tieghem's Bacillus amylobacter and is then 

 split up into equal volumes of methane and carbon dioxide, accord- 

 ing to the equations 



= C 6 H 12 O 6 . 

 C 6 H 12 6 = 



He found the ratio different when sulphates (gypsum, <fcc.) or 

 ferric salts were present in the water or mud. In such case the 

 nascent methane, by its reducing action on these salts, converts 

 them into carbonates, sulphuretted hydrogen being liberated, accord- 

 ing to the equation 



CaS0 4 + CH 4 = CaC0 3 + H 2 S + H 2 0. 



The sulphuretted hydrogen is, then, under natural conditions, 

 acted upon by the sulphur bacteria which are always present in 

 such waters. This will be dealt with in Chapter xxxv. 



The importance of cellulose fermentation in the physiology of 

 nutrition (especially of cattle) must also be briefly adverted to. 

 The opinion long held by Emil Wolff, that the vegetable fibres 

 consumed with the food pass out of the alimentary canal unaltered, 

 was contradicted, in the case of ruminants, as far back as 1854, by 

 Haubner, who showed that even in the case of sawdust and paper- 

 pulp mixed with the fodder, only a portion (less than half) was 

 expelled again in the excrement, a fact confirmed by the exhaustive 

 researches of Henneberg and Stohmann. The difference between 

 the amounts of cellulose (crude fibre) taken in and rejected was 

 highest in the case of ruminants (up to 75 per cent.), being only 

 50 per cent, at most in horses, and still less in the human subject 

 and in swine. In carnivora (dogs), on the other hand, no such 

 difference could be detected. The quantities of cellulose thus 

 disappearing in digestion were, it was thought, digested, and were 

 regarded as approximately equivalent in nutritive value to the 

 soluble carbohydrates. A number of animal physiologists main- 

 tained that the cellulose was dissolved by an intestinal enzyme, 

 which, however, was sought for in vain. The earliest reliable 

 determinations on this point were made by H. TAPPEINER (I.) in 



