418 SECTIONAL TRANSACTIONS— M. 



the manuring, and poppies rose to over 250 millions per acre with complete fertilisers. 

 Some weeds are unafiected by the manuring, but others are definitely encouraged or 

 discouraged by various combinations of manures. The response of different species 

 of weeds to two years' consecutive fallow varies considerably. Some species, as 

 Alopecurus agresiis (black bent) and Polygonum aviculare (knotgrass) are almost 

 eliminated, whereas Papaver spp. (poppy) is only reduced to about 46 per cent, of the 

 original number, and a few others are very little affected. This is due to the varying 

 lengths of time that buried seeds naturally remain dormant even when they are 

 brought under conditions favourable for germination. Consequently, fallowing 

 might prove economic for heavy infestations of some weeds as black bent, but for poppy 

 and others the clearance would not be good enough to justify the expense. 



The times of maximum germination of weed seeds show a marked periodicity, most 

 species having a strong preference for starting into growth in either the autumn or 

 the winter quarter, comparatively few coming up in the spring or summer. Land 

 should therefore be broken up as soon as possible after harvest and kept on the move 

 in order to obtain the maximum reduction of buried weed seeds, whether the land is 

 to be cropped or kept under fallow. The experiments are being continued to determine 

 the after-effects of fallowing when the field is back under crop. 



PRETORIA. 

 Wednesday, July 31. 



Discussion on Methods op Soil Investigations in Field and 

 Laboratory. 



(a) Soil Classification : 



(i) Field Methods. — Prof. G. W. Robinson (Agricultural Soils), Mr. 

 A. Stead, Mr. C. G. T. Morison (Forest Soils). 



Prof. G. W. Robinson. — Soil Surveys and Soil Classification. 



Some kind of classification must accompany everj' type of soil survey. Consider- 

 able attention has been devoted to the formulation of systems of classification which 

 shall embrace soils from aU parts of the world. Such systems have been proposed by 

 Glinka, Lang, Bamann and others. Marbut has recently outlined a scheme which has 

 the advantage that it is sufficiently elastic to include new types yet to be discovered. 



In the survey of a new country, however, the place of its soils in a world system of 

 classification does not call for immediate attention. The most important task is to 

 recognise what are the significant types, and to arrive at a classification which is valid 

 for the country in question. In arriving at this classification, two considerations 

 should be borne in mind. Firstly, the classification should be based on the soil itself 

 and not on the factors influencing soU formation. Where, however, a close corre- 

 spondence can be shown between the character of the soil and one of these factors, 

 e.g. geology, or topography, such factors may reasonably be used as aids in mapping. 

 Secondly, the unit of study should be the complete soil profile and not simply the 

 surface soil. 



Generally speaking, the main types of soil in any country can be recognised by 

 field observations. Laboratory examination serves to give quantitative significance to 

 type descriptions and may also help in the final subdivision of areas marked for 

 intensive survey. 



When the soils of a country have been properly arranged into types among them- 

 selves, it will be possible to consider their place in a world classification. The 

 accumulation of information as to new soil types must necessarily modify schemes of 

 classification already proposed. 



Mr. C. G. T. MoBiSON. — Records required in surveys of forest soils — extensive and 

 intensive surveys. Vegetation characters — permanence and depth of leaf litter — 

 depth of profile observations— the development of aerial surveys. 



(ii) Laboratory Methods : Chemical Analyses. — Mr. A. Stead, Dr. 

 W. G. Ogg, and Dr. M. S. du Toit. 



