LUCKRXK IX THE TRANSVAAL. 41 



per Ib. this means a erst of 10s. per acre more for broadcasting 

 than for drilling. In addition to this, drilled lucerne sometimes 

 uives a much larger yield per acre than that which he.s been 

 broadcast, so that the advantage is all in favour of drilling. 



On dry lands the drills should be 9 inches to 12 inches apart, 

 as the plants do not spread much, but under irrigation it, is desir- 

 able to space the rows to 15 or 18 inches on account of the tendency 

 of the crowns to branch laterally. 



Weeds are fatal to good stands of lucerne ; it is therefore of 

 the greatest importance that the seed used shall be absolutely clean 

 .and free from weed seeds. Where so much seed is imported frcm 

 abrcad there is danger of introducing new and noxious weeds. 

 There is also a tendency to dump on the South African market any 

 old seed of po; r germinating power which will not sell readily in 

 other countries ; where such high prices are paid for imported 

 seed, it is wasteful to buy any but the best. Lucerne seed of 90 

 per cent, purity and germination, for which Is. per Ib. is paid, 

 actually costs Is. 1-Jd. foi every pound of the seed that grows. 

 "Where only 9d. per Ib. is paid, but the combined germination and 

 {unity do not exceed 50 per cent., the actual price is, of course, 

 Is. Gel. for every pound of the seed that grows. It is therefore 

 much cheaper to buy good seed, of high purity and germination, 

 at Is. per Ib., than poor seed of low germination, at 9d. per Ib. 

 But it is not inly u qr.estkn of the actual cost of the seed; pc or 

 seed gives a poor stand, and a poor stand gives a poor crop. V\ ith 

 a permanent crop like lucerne the stand should be as perfect as it 

 is possible to make it, in order to secure full returns. 



On irrigated land which is absolutely clean of v\eeds it may 

 sometimes do to sow in early spring, thus getting the stand estab- 

 lished before the weeds have much chance to get ahead. Many of 

 our weeds wait, till summer is well advanced before starting vigor- 

 ius gr will. This course is, however, very risky, for there is danger 

 from late spring i'r, ;sts, and there is sometimes as great a tendency 

 for the lucerne plants not to "come away " quickly as there is for 

 the weeds to grow slowly ; the same influences afiect both; if the 

 IIK erne pushes ahei d, the weeds are apt to do so too. 



Our experiments carried on during some five years, in every 

 dis rict in the Transvaal, on practically every kind of soil, and in 

 almost every month of the year, show that sowings made during 

 the latter part of summer and early nitumn, from the middle of 

 .January to the middle of 3,!arch, give far the best stands ; by this 

 time the weeds are harmless, having passed their prime and 

 xhaustiii themselves with seed. The lucerne seedlings continue 

 Crowing as long as there is any moisture available, until late on in 

 the winter, and are thus well prepared to rush into vigorous growth 

 wi;h I ho first warm weather of spring. This does not mean, how- 

 ever, that one can even then afford to sow lucerne in foul land. A 

 certain proportion of weeds will appear in the cleanest of land, but, 



