26 Seed Wheat. 



SECOND TRIAL. 



The second set of trials in connection with the relative seed value of large, medium 

 sized, and small grains of wheat was undertaken in 1895. The tabulated results of these 

 trials are given below. (See page 27.) 



It will be seen that the experiments include 117 trial rows containing twenty-three 

 varieties of wheat, representative of all the types that are largely grown in the State. 

 The trials cover about seventy different comparisons between equal numbers of seeds of 

 varying grades or sizes, the comparison being between the yields of grain and straw from 

 larger grains on the one side and smaller grains on the other. 



The season was an average season, and the land was typical River ina wheat land with 

 perhaps a tendency to vary more than usual towards freshly-decomposed granite. The 

 rows were placed side by side, and occupied, with the other rows among which they were 

 distributed, a space 2 chains wide and 3 chains long. The land sloped gently toward 

 the west, and improved in quality from top to bottom. The "small-seed" rows were in 

 all cases on the down-hill side, so that whatever difference in fertility there was from 

 the gradual change for the better at the lower end of the plot was in favour of the small 

 seeds. This small balance in favour of the small seeds would be trifling in amount, as 

 the rows compared were never more than 2 links apart, and I have, therefore, disregarded 

 it. There is no way of satisfactorily estimating its amount, and it is, accordingly, 

 necessary to content oneself with the general statement of its existence and the 

 insignificance of its amount. 



It must be borne in mind, however, that the various pairs of rows were, as a rule, not 

 adjacent to each other, so that the reader must guard against the impression that in the 

 case of any variety he is reading the records of a succession of rows. In the table the 

 various rows are numbered in succession for convenience of reference ; but the actual 

 growth was in a different order. There is no chance of error in reading if the reader 

 simply notes what comparisons are to be made, and, as before pointed out, this is only 

 permissible in the case of two rows growing next each other. Sometimes the middle row 

 of three is compared with each of the two adjacent rows. 



The rows were cared for by hand during the entire season. There was no cultivation 

 beyond that necessary to keep down the weeds ; but this was an amount that must be 

 taken into consideration in interpreting the results. The weeds were pulled by hand, or 

 were cut out with a hand hoe. There was no desire to cultivate, but simply a desire to 

 give the trial rows, as nearly as possible, identical conditions. If the weeds would have 

 grown uniformly, they would have been allowed to grow. As they would not grow 

 uniformly, but would grow more in one place than another, and thus rob one row 

 more than another, it was thought best to keep them down with as little disturbance to 

 the soil as possible. 



This removal of the weeds is, it seems to me, decidedly in favour of the smallest seeds, 

 as the trials show that nearly if not quite all adverse conditions are more severe on the 

 smallest seeds. Hence relieving the smallest seeds from the competition of weeds was a 

 greater favour to them than to the largest seeds. I do not consider that this matter 

 was of much importance in comparisons of adjacent rows from seed above the 250 grade 

 or thereabouts, but I think that the tailings and the 200 grade benefited considerably by 

 this treatment as compared with the larger grades. 



If there be objection to assuming that the method adopted would really favour the 

 smaller seed, there can hardly be any objection to assuming that at least there was no 

 favour shown the large seed, unless, indeed, it be assumed that culture favours large 

 seed more than it does small seed, an assumption so far removed from probability that 

 I think it may be safely disregarded. 



If the method followed was neutral so far as relative yield is concerned, then in 

 calculating how much a farmer can afford to spend in improving his seed we have a 

 simple proportion. If, for instance, the yields here given are twice those the farmer is 

 accustomed to get. then the extra yield here shown to be due to the use of large plump seed 

 has only to be divided by two to give the figures upon which the farmer may proceed. 



An examination of the tables will at once show that under the conditions of the trials 

 there was inferior germination and growth of the smallest seeds, or perhaps it would be 

 better to confine the statement to that part which relates to growth. The table pretends 

 to show nothing as to the precise amount of germination that took place, the statistics 

 being confined to the number of plants that actually grew and presented themselves at 

 harvest-time in a condition in which it was possible to cut and weigh them. It may be 

 as well to state, however, that the plants missing at harvest-time were almost wholly 

 instances of failure to properly germinate, at any rate in the cases of the smallest grains. 

 If these lost plants appeared above ground, they were weak seedlings that could not 

 compete with ordinary conditions, and so succumbed. The care exercised was such that 



