43 



seed of the species of which the parcel purports to consist, in terms of a 

 percentage of the total weight, and it is this figure which the consvuner is 

 entitled to Imow. Members of the seed trade are perfectly aware of the 

 significance of the terms of the conventional report, but there is widespread 

 ignorance amongst consumers as to the actual meaning of the figures 

 therein provided. 



There is little virtue in the adoption of the pedantic attitude that 

 reports are only reports of experiments and should be understood as 

 such. Especially when a Station is a " Control " Station, it must be 

 prepared to substantiate its findings as facts. 



It is, therefore, here proposed that reports should be framed with 

 a view to suppressing the figure for germination, — -which, being based 

 on an experiment carried out on the selected pure seed means nothing 

 in itself and is only of significance when considered with the ascertained 

 purity, — and introducing a form of report in terms of percentage of pure 

 germinating seed and percentage of impurities. 



I. One advantage of such a procedure is that the report would show 

 at first glance the actual percentage weight, as nearly as can be calculated, 

 of the live seed of the kind supplied, in contra-distinction to the figure for 

 germination which has a certain falsity. 



Individual seeds of the same species in the same sample differ in size, 

 weight and condition. 



The selection of seeds of different weight and germinating values in 

 their due proportions from any sample is assumed to be subject only to 

 the limits of mathematical variation. This has been examined by 

 Rodewald and by Stevens, who have shown that the probable error in 

 selection is not, in general, large. 



But, if a sample consists of seeds which can be graded into different 

 weight sizes, and if there is any difference in the capacity for germination 

 of ■Qie various portions of the sample thus graded, then the proposition 

 that the estimate of the percentage of pure germinating seed is a true 

 weight percentage is not truly valid. 



Critical examples which might seem to invalidate the proposition fall 

 into two classes : — ■ 



(1) "Seeds," which are normally at Seed Testing Stations 

 excluded from the germination test, i.e. — - 



Broken seeds of leguminosse. 

 Damaged seeds. 



" Deaf " or " light " seeds of grasses. 

 Seedless clusters of Beta. 



(2) " Seeds," the germination value of which can only be satis- 

 factorily established by referelice to a germinating test, i.e. : — 



Small mature seeds. 

 Shrivelled or unripe seeds. 

 Shelled caryopses. 

 Weakly germinating seeds. 



Broken and damaged Seeds. — The difference in weight between 1,000 

 broken seeds of Red Clover as usually excluded in analysis and 1,000 whole 

 seeds varies greatly in different samples — ^from 2 per cent, to 20 per cent, 

 of the weight of whole seeds. But some of the broken seeds, as defined 

 in Seed Testing literature and as customarily excluded from the germina- 

 tion test, are capable of germination if germinatio|i , lae determined 

 according to accepted rules, viz., that all seedlings which possess two 

 cotyledons attached and which have, at the end of 10 days, thrown a 

 secondary or an adventitious rootlet, are to be reckoned as germinated. 



Broken seeds should, however, only be considered as equivalent to 

 inert matter when it is absolutely certain that the cotyledons are separated 

 and that the plumule has been lost. 



Ab a general rule, it is waste time on the part of Seed Testing Station 

 assistants to try to identify " broken " seeds, unless the term be reserved 

 for chips which can be segregated without special examinations. In 



