376 



Annual Report for 1911 of the Botanist. 



l)ut evidently had made no attempt to clean the seed after 

 threshing. To sow such rubbish is a highly efficient method 

 of obtaining land overrun with docks. The second sample — 

 sainfoin — was badly contaminated with its usual impurity, 

 burnet, and moreover failed to germinate 20 per cent. To 

 secure a full plant from such seed would have entailed sowing 

 nearly ten times the usual quantity per acre. A third ex- 

 ceedingly bad sample was sent too late in the season for the 

 germinating tests to be of any assistance. This was a small 

 quantity of a new wheat which had failed to make a satisfactory 

 plant, though another variety sown in the same field at the 

 same time had grown perfectly. Under laboratory conditions 

 it germinated 22 per cent, as against the 5 per cent, the 

 Member estimated that it grew in the open field. 



The following table shows the average purity and ger- 

 minating capacity of those kinds of seeds submitted in suffi- 

 ciently large numbers to give reasonably representative 

 results : — 



The figures indicate that the agricultural seeds now supplied 

 are satisfactor5^ I bSlieve, though, that most of the seeds 

 tested have been obtained from the larger seed merchants, and 

 that the average results would be distinctly lower if samples 

 representing the seeds supplied by small local dealers and corn 

 merchants were available for analysis. This applies more 

 especially to gi'ass and clover seeds, which require elaborate 

 machinery to clean them effectively. The English and Con- 

 tinental samples of clover seeds tested have proved to be cleaner 

 than in former years, and dodder was completely absent from 

 all but two samples. Chilian clover seed, on the other hand, 

 generally contained the large seeds of Chilian dodder, which 

 appear to be difficult to separate completely. The average 

 content of hard seed in the four clovers in the above list was 

 5'(S per cent. 



