II Oct., 1909.] Seed Tests. (>t^t, 



Beta vulgaris (-AJani^t^l Wur;cel). — 100 fruits contained 220 seeds of 

 which 140 germinated; 8 per cent, fruits contained i seed; 72 per cent., 

 2 seeds; 12 per cent., 3 seeds; and 8 per cent., 4 seeds. Of these, 17 

 fruits produced i seedling; 17 fruits, 2 seedlings; 5 fruits, 3 seedlings; 

 and 2 fruits, 4 seedlings. 



Brassica oleracea (Cabbage). — Sample pure, but seeds very unevenly 

 ripened and varying in size from i to over 2 m.m. 



Brassica rapa (Rape). — Sample jjure, but seeds \-aried from 1.5 to 2.5 

 m.m. in diameter. 



Cannabis sativa (Hemp). — Seed clean and free from weed seeds, but 

 gave low germination result. 



Cucnmis Melo (Musk Melon). — Excellent sample. 



Cnmintim C yminiim (Cumin). — A ver\- low germination value. 



Cynodon Dactvlnn. — The result speaks for itself. This seed was 

 actuallv bought and planted by at least one pri\ate individual, but, of 

 course, without any result. 



Cynosiirus cristatus (Crested Dog's Tail). Germinated verv irre- 

 gularlv ; 22 per cent, came up quickly; the remainder required from 3 — 5 

 weeks in all. Manv of these would fail in the soil. 



Dactylis glomerata (Cocksfoot).— Containing a few weed seeds, chiefly 

 a Composite {Crefis fatida,. Stinking Crepis). Verv slow in germinating, 

 and were kept in germinating chamber over 5 weeks. 



Linum usitatissimum (LinseedV — Contained a number of weed .seeds, 

 includnig Convolvulus (Bindweed), Melilotus, Charlock and other 

 Crucifers. 



Loliuin fercnne (Rye grass). — Being from New Zealand, this seed had 

 only been subjected to a short sea journey. 



Pastinaca sativa (Parsnip). — Clean sample of seed, but containing about 

 16 ix*r cent, unfertile seeds. Kept in germination chamber for seven 

 weeks, when 13 seeds still remained ungerm.inated and undecayed. These 

 would be practically valueless as seed, owing to the excessive delay in 

 their germination, even if ultimately they formed seedlings. 



P/ialaris canariensis (Canary seed). — Contained a number of weed 

 seeds, including Melilotus, two species of Polygonum, Charlock (Brassica 

 Sinafistruni) Johns-ton Grass (Andropogon halepensis), Ragweed 

 {Ambrosia art c i)ii si (1; folia) and Purple Cockle {Lychnis Githago). 



Rheum (Rhul)arb). — Above the germination vabae. The only certain 

 local sample tested on the list. 



Sesamum indiciim (Sesame). — Possibly intended for grain, not for seed. 



Solarium Lycopersicum (Tomato). — Possibly a locally grown sample. 



Tri folium Alexandrijvim (Egyptian Clover). — Contained a small 

 number of weed seeds, chieflv two forms of Composite and a species of 

 Dock. 



Vigna sinensis (Cln\ Cow Peas). — 14 per cent, were hard seeds and 

 required treatment with strong sulphuric acid for one hour to make them 

 swell. Not a verv good sample, as seeds unevenly ripened and 5 per cent, 

 damaged. 



Zea Mais (Pop Com). — 13 per cent. bad. Xo weed seeds present, but 

 the grains were nibbled by mice and pierced bv weevils. 



From the foregoing list it can be seen that the largest number of 

 weeds were found in the seeds from India anrl Turkey. In only seven 

 ca.ses did the germination value exceed or come up to the limit usually 

 set for good seed. In five cases, the germination value was one-third of 

 what it should be, or even less. In a total of .se\enteen cases, the 



