272 DTEECT SOWING. ' 



soil or by using soil from excavations made alongside. In the 

 latter case, the excavations serve as drains and thus contribute 

 powerfully towards root-aeration. In localities where the soil, 

 although water-logged during the rains, becomes very dry during 

 the re3t of the growing season, the excavations will catch from 

 occasional showers enough water to keep the plants alive and 

 vigorous. The excavation may consist of a single square pit on 

 one side of each mound, or of two pits, one on each of two opposite 

 sides of the mound, or, lastly, of a sort of continuous ditch or moat 

 surrounding the mound, according to the degree of swampiness to bo 

 corrected. The excavations help to reduce cost by making it 

 possible to diminish the height and generally the size of the 

 mounds. 



Sowing on mounds constitutes far excellence the system to be 

 followed in very wet soils, where the mounds would not only be far 

 more effective than ridges, but also cost very much less. The 

 height at which the seedlings would start would render them safe 

 against invasion or suppression by powerful weeds. 



ARTICLE 14. 

 Combined patch, pit and mound sowing. 



As in the case of continuous cultivated lines, four different com- 

 binations are here also theoretically possible: — (1) hollow and 

 mound, (2) patch and pit, (3) patch and mound, and (4) patch, pit 

 and mound. 



The only useful form of these combinations is the one already des- 

 cribed under the preceding Article, viz, a mound rising out of a 

 hollow. Here the sides of the mound and the surrounding hollow 

 would be sown with a moisture-loving species capable of serving as 

 a nurse to the more valuable species sown on the top of the mound. 



The second, third and fourth combinations are, for the present at 

 least, of no practical utility for us. 



SECTION V. 

 Quantity of seed to sow. 



We are now in a position to appreciate the quantity of seed to 

 sow in any given case. It will depend — 



(i) On tlie quality of the seeds, i.e., the number per cent, that 

 are expected to germinate, the average time in which they are 

 likely to complete their germination, and the vigour of the seedlings 

 that it is believbd they will produce. 



(ii^ On the method of soic'mg employed. — Thus complete sowino- 

 without tilth will require most seed ; then follow, in descending 



