II Nov., 19 1 2.] General Notes. 665 



marlcfd degree — they contained much more nitrates and ammonia 

 early in the season. No reason could be given Tor this. Botanically, 

 there was very little difference in the types of herbage. The feeding 

 value of the pasture was thus largely independent of its botanical 00m- 

 position. Habit of growth appeared the important thing. On the good 

 land, with plenty of nitrates and ammonia, the grasses were leafy, with 

 little tendency to form heads; on the bad fields the herbage was s'temmy, 

 and flower heads came early and abundantly. Altogether the investi- 

 gations indicate that the bacterial activity of pasture lands is of first- 

 rate importance. They suggest that the application of lime (a germ 

 tonic) would often bring marked improvement. Lime increases the 

 natural supply of nitrates. Incidentally it helps clovers in another 

 way. On some soils phosphates are required in addition to lime. It 

 is by experimenting along these lines that the right means of improving 

 poor pasture can be determined. 



HERD TESTING— 



Six years ago Mr. Burgess, a New Zealand farm.er, started testing his 

 dairy herd, and a lucid account of his experience appears in the ]ournal 

 of Agriculture of the Dominion. In the first season of testing the 

 cows averaged 198 lbs. of butter fat. As a result of always weeding 

 out the worst cows, the average for the second season was 222 lbs. ; 

 for the third season 234 lbs; for the fourth, 241 lbs.; in the fifth 

 season the yield was affected by the dry summer, and consequent scarcity 

 of feed; in the sixth season the average was 261 lbs. Six year's testing 

 thus increased the yield by 63 lbs. butter fat per cow. The cost of 

 testing is calculated at 8s. iid. per cow, including cost of apparatus 

 and the farmer's own time at is. per hour. He estimates that improve- 

 ment will continue, because in the period under review an exceptional 

 number of heifers ^^as introduced to replace culls. The herd was of 

 mixed breeding, but at two dates a pure-bred Ayrshire bull of good 

 milking strain was obtained. The writer would prefer pure breeds to 

 work on as these perpetuate their qualities with greater certainty, but 

 points out the difficulty of obtaining this class of stock with good records 

 well authenticated. The scheme for officially testing pure herds recently 

 introduced by this Department will furnish a guarantee to those purchasing 

 pure-bred animals with the object of grading up their herds. 



DRAINAGE LOSSES— 



Phosphoric acid, potash, lime, and nitrogen are the soil constituents 

 which the farmer wi.'-hes to conser\'e, and each of them stands in a 

 different relation to drainage losses. A discussion of the subject and 

 the analysis of drainage waters from the fields of five farms which 

 appeared recently in the Illustrated Landw. Zeit, which serve to draw 

 attention to the subject. The results were typical. There was no 

 phosphoric acid lost by drainage. The loss of potash was not serious, 

 but there was considerable loss of nitrogen (11.8 parts per million) 

 almost wholly as nitrates. There was no loss of nitrogen as ammonia. 

 The most serious loss was in lime, amounting to 215 parts per million. 

 In nearly all cases lime will be the chief constituent in drainage waters. 

 The loss is greater on cultivated land than on pasture, and most of the 

 manures in use tend to increase it. The waste is inevitable, and in 

 the long run must be made up by fresh applications of lime at the surface. 



