80 



growth of the plant to nearly ten per cent, of water on the average 

 in the top three feet of soil, and the difference was felt to four feet 

 deep. Now barley is only a surface-rooted crop, just like a large pro- 

 portion of our weeds, and the illustration shows what great evil even 

 surface-rooted plants may do to the water condition in the lower layers 

 of the soil. It should therefore be laid down as an absolute rule that 

 weeds growing among the tea after the end of the rains are dangerous, 

 and so far as is possible should be avoided. 



The same applies to green manuring crops put in in October and 

 November. Much " Mali Kalai " has at one time or another been 

 grown then in imitation of native culture, but one can hardly condemn 

 the system enough, because it removes soil moisture just when it is 

 required and in addition, as a matter of fact, very little green matter 

 results at that time of year from its growth. Green manures should be 

 sown, on principle, when the land is increasing in moisture content, 

 and hence in April or May, when rains are expected in greater and 

 greater quantity. They then- grow far more luxuriantly, give more 

 green matter for hoeing in, and do not rob the tea plants of the water 

 which they are needing. 



EFFECT OF TOO WIDE DRAINS. 



But the needed water is also often taken away from the soil by 

 the existence of very wide drains. On one garden I have visited 

 the drains form by far the best riding roads through the property 

 in the dry weather, and in many others they are almost as wide. 

 In some cases, especially where a sandy soil rests on a clay subsoil, 

 there seems little help for this, as they continually fall in until this 

 soil takes the natural angle of repose, but one could imagine, never- 

 theless, no more effective means of drying up the land, and so, to a 

 certain extent, drying out the roots. Such drains are, if they can be 

 helped, a mistake and a serious one. Drains should be as narrow as 

 possible not more than the width of a kodalli wide if the texture 

 of the soil allows of this and they should, where it can be done, be 

 so banked up at the top that the sun, with its drying action, 

 penetrates the drain to as small an extent as can be arranged. 



IRRIGATION FOR TEA. 



There are, however, cases where all these means do not suffice 

 to retain sufficient water in the soil for the best results. Such a 

 case occurs on the bank of highly fertile micaceous sand which 



