EXPERIMENTS OX TUENIPS. 321 



The addition of f cwt. nitrate of soda gives a mean increase of 

 2'3 cwt. dry matter, so that about one-third of the increase in 

 weight of bulbs (40 cwt.) shown in Table VIII. is water, and 

 the additional turnips grown by the use of the nitrate along 

 with dung will cost about 7s. per ton. 



The next point worthy of note is the comparative composition 

 of turnips grown with dissolved and undissolved phosphates. 

 Taking the dunged plots first, the mean dry matter in plots 2, 3, 

 4, and 20, which got superphosphate, is 8*6 per cent., and ash in 

 dry matter 7'65 per cent., and the meau results of the correspond- 

 ing plots which got ground coprolites are practically the same, 

 namely, dry matter 8*65 per cent., and ash 7"5 per cent. On the 

 undunged plots, on the other hand, superphosphate gives a rather 

 better result than ground mineral. The mean of the dry matter 

 iu the superphosphate plots (23, 24, 25, 13) is 9'1 per cent, 

 while in the corresponding coprolite plots (9, 10, 11, 12), it is 

 8'8 per cent., and the percentage of ash in the dry matter is 

 higher in the latter — 6'8 — against 6*5 in the superphosphate 

 plots. The phosphoric acid was determined in these ashes, and 

 gave in the bulbs grown with superphosphate ll'ol per cent, of 

 the ash, and in those grown with coprolites lO'Oo per cent., so 

 that, although the percentage of ash is higher with coprolites, 

 more phosphoric acid was absorbed by the turnips when super- 

 phosphate was used. The mean dry matter per acre of the eight 

 superphosphate plots given in Table XIII. is 47"9 cwt., while in 

 the eight corresponding coprolite plots it is 45 "9 cwt., showing 

 a superiority of nearly 5 per cent, in favour of superphosphate ; 

 whereas there was only 3 per cent, of difference in the actual 

 weights of the turnips on the plots. Making allowance for 

 unavoidable errors in sampling, it may be fairly inferred that the 

 bulljs grown with superphosphate this season were at least as 

 good as those manured with coprolites. 



To compare the composition of the turnips on the dunged and 

 undunged plots, the analytical results can be inserted in Table 

 X IV. Averaging the numbers, the following results are obtained: 

 — Mean of ten plots with dung, dry matter 8"6 percent, and 51-5 

 cwt. per acre, ash in dry matter 7 6 per cent, and 437 lbs. per 

 acre. Mean of ten plots without dung, dry matter 8*95 percent, 

 and 432 cwt. per acre, ash in dry matter 67 per cent, and 323 

 lbs. i)er acre. There were 6 tons or 25 per cent, more bulbs on 

 the dunged plots than on those that got artificials only, but the 

 superiority is decreased by 5 per cent, if the dry matter of each 

 be compared, or 1 ton of the increase is water. One-third 

 more ash constituents are removed from the soil on the dunged 

 than on the undunged plots. Indeed, the difl'erence in the 

 percentage of ash on these plots is very noticeable, as it is large, 

 and occurs in all the plots compared. 



X 



