TRANSACTIONS OF SECTION K, 851 



production of marsh gas in swamps, the question of the digestion of cellulose in 

 herbivorous animals, the manufacture of ensilage, and the processes of ' shorten- 

 ing ' of manure ; and it is clear they have much to do with the destruction of 

 paper, &c., in sewers and refuse-pits. Moreover, their further investigation pro- 

 mises a rich harvest of results in explanation of the rotting of stored tubers, certain 

 diseases of plants, and several theoretical questions concerning anaerobism, butyric 

 fermentation, and, possibly, that extremely difficult question on which Mr. Gardiner 

 has done such excellent work, the nature of the various celluloses and constituents 

 of the cell-wall. 



I now turn to the subject of fungus epidemics, of world-wide interest, if only 

 because the annual losses to agriculture due to epidemic diseases of plants amount 

 to millions of pounds sterling. 



'The history of wheat-rust can be traced to Genesis, and at least five references 

 to it exist in the Old Testament. The Greelis were familiar with it, and the 

 Romans had a special deity and ceremonies devoted to it. References can be 

 given to it in old Norman times, and Shakespeare can be quoted as acquainted 

 with it. 



According to Loverdo, a law existed in Rouen in 1660, authorising the pulling 

 up of barberry bushes as in some mysterious way connected with rust, and in 

 1755 the celebrated Massachusetts law was promulgated. Eriksson refers to an 

 English farmer destroying his neighbour's barberry in 1720. 



The words Robigo, Rubigo, Rouille, Rurjgine, Rtifus, and Rust comprise a his- 

 tory in themselves, into which, however, we have not time to go, and there are 

 many fascinating points in the history of wheat-rust which must be passed over. 



Felice Fontana in 1767 probably made the first scientific investigation of rust; 

 he distinguished the uredo- and puccinia-stages under other names, and eveu 

 thought of them as rootless plants exhausting the wheat ; in this, and his convic- 

 tion that no remedy was possible until a careful study of all phases of the disease 

 had been made, he was far ahead of his times. 



Jethro TuU, Marshall, and Withering are the most conspicuous English names 

 in connection with this question and period, and Marshall in 1781-84 experimented 

 intelligently with barberry and wheat inter-planted. 



Persoon in 1797 gave the name Puccinia graminis to the fungus. In 1805 Sir 

 Joseph Banks described it, and suggested that the germs entered the stomata : he 

 also warned farmers against the use of rusted litter, and made important experi- 

 ments on the sowing of rusted wheat-grains. 



A great discussion on the barberry question followed, in which Banks, De 

 Candolle, Windt, Fries, and others took part. Fries particularly insisting on the 

 diiference between ^cidium berberidis — a name conferred by Gmelin in 1791 — 

 and Puccinia gramitiis. 



De Oandoile had also distinguished Uredo rubigo-vera in 1815, and Schmidt 

 soon after described a third wheat^rust — Uredo glumarum. 



Matters were at about this stage when Tulasne confirmed the statement of 

 Henslow — one of my predecessors in Cambridge — that the uredo- and puccinia- 

 stages really belong to the same fungus, and are not, as Unger asserted, mixed 

 species. 



Then came De Bary and his classical investigation of the whole question in 

 1860-G4. He proved that the sporidia of some Uredinefe (e.g., Coleosporii/»t) will 

 not infect the plant which bears the spores, and that the secidia of certain other 

 forms are only stages in the life-history of species of Uromyces and Puccinia. 



In 1864 be Bary attacked the question of wheat rust, and by means of 

 numerous sowings of the teleutospores on barberry proved beyond doubt that they 

 bring about its infection. 



But De Bary did more. For the first time in history he saw the entrance of 

 the infecting tube and the beginning of its growth in the tissues. In 1865 he 

 demonstrated in the same faultless way the infection of the cereal by means of the 

 aecidio-spores, and showed that P. rubigo-vera alternates on Boraginea; as yEc. 

 asperifolii, while P. coronata, separated by Corda in 1837, does the same as yEc, 

 Mhamni on Rhamnics, 



?I2 



