W. J. DowsoN 89 



much more slowly in the dry season. Both C. arabica and C. nandiensis 

 are subject to a number of fungous and insect enemies. Meteorological 

 conditions play a most important part in the severity of attack of both 

 fungi and insects, particularly so in the coffee leaf disease due to the 

 rust Hemileia mstatrix B. & Br. 



So far as the writer is aware, coffee cultivation in the eastern hemi- 

 sphere, including Africa, has always been intimately connected with 

 that of HemUeia, and has usually resulted in the ascendancy of the 

 parasite sooner or later. This does not necessarily mean that the coffee 

 trees are threatened with destruction as was the case in Ceylon, but it 

 does mean that Hemileia is slowly spreading in all those countries in 

 which coffee is grown, with the present exception of the highlands of 

 East Africa. 



It is hoped to deal with the subject of coffee leaf disease in Central 

 and Eastern Africa in more detail in a further paper. For the present 

 purpose, some observations of the disease as it is now found in the 

 highlands of Kenya Colony will suffice. 



The first observation of importance is that only once in Africa and in 

 Ceylon have the teleutospores of the parasite been found. It is a curious 

 and as yet an unexplained fact that since Marshall Ward worked out 

 the life-history of the fungus in Ceylon, and a German observer reported 

 the presence of teleutospores on some African specimens of the disease 

 not long afterwards, these spores have never been observed since. The 

 likelihood, therefore, of the existence of an aecidial stage on some other 

 plant is not very great. Coffee leaf disease, like many other rusts, is 

 propagated in the countries in which it is found by the uredospores only. 



The second observation to be recorded is, that the first attack of 

 Hemileia is undoubtedly the most severe; subsequent attacks, other 

 things being equal, are less marked in intensity. Nearly all the trees are 

 badly infected, but in well-kept plantations only a small percentage of 

 the leaves actually fall, although the lives of the others are considerably 

 curtailed. Entire defohation never takes place, and subsequent attacks 

 are less severe, that is to say, not so many pustules of uredo-sori are 

 formed and not so many leaves are infected. That the general health of 

 the trees has much to do wdth the severity of the attack is obvious when 

 an ill-kept plantation is compared with others better cared for in the 

 vicinity at any season of the year. The initial preparation of the ground, 

 the proper planting of the seedlings, pruning and the amount of berries 

 the trees are allowed to carry, are all factors which influence the resist- 

 ance of the host. 



