124 



(d) Characteristics of the seedlings as to colour, presence or 

 absence of hairs or other types of indumentum, shape and size of 

 the cotyledons, hypoootyl, and the first leaves of the seedlings. 



Under the above items some kinds of seeds and fruits can be very 

 <?asily distinguished. Thus, in Brassioa, Raphanus, Daucus, Allium, 

 "Capsicum, Lactuca, Spinacia, Beta, etc., varietal and form characters can 

 be distinguished fairly well from the seeds or fruits alone. It is scarcely 

 necessary to mention here that the. various cereal and leguminous crops 

 present a great wealth of varieties, but with these, as with representatives 

 of those genera mentioned above, varieties can be easily distinguished 

 from the seed characters alone. A prime essential, . however, for the 

 accurate identification of garden varieties of seeds is an intimate luiowledge 

 of the different forms, a knowledge that can only be gained through inten- 

 sive study and long experience. 



In Japan, as elsewhere, agricultural seeds have been but slightly 

 investigated. This is all the more curious when we reflect that agriculture 

 is the absolute basis of our civihsation, and anything that will forward 

 agriculture will, of course, have its effect, direct or indirect, on modem 

 civihsation. My experience leads me to believe that scientific investiga- 

 tions, such as those that I have carried on for a number of years on the 

 identification and standardization of agricultural and garden seeds, are not 

 •only of very great interest and usefulness, but are definitely practicable. 

 One by one the various economic species are being taken up and their many 

 varieties and forms are being intensively studied. Since Harz's book, 

 published in 1885, very little has appeared on this subject in the scientific 

 .and agricultural literature of Europe and America. 



Siunmarizing, I would state that in seed testing, the question of the 

 percentage of germination and the question of contamination by weed seeds 

 are by no means the most important considerations, for strains must be 

 tested as to nomenclature ; that is, that the seeds are true to the type of the 

 variety or form under which they are sold. Until very recently, no investi- 

 gations on this important subject have been made ; it is, however, clearly a 

 most important part of seed testing. Such investigations should be greatly 

 extended and adequately supported as a vitally important branch of 

 agricultural science. Research should not be confined to the seeds 

 themselves, but should be extended to the seedlings both from the botanic 

 And agronomic viewpoint. The major subjects for observation and 

 ■comparison should be : — 



(a) External characters. 



(6) Size, weight, specific gravitj'. 



(c) Anatomical structure. 



(d) Seedling characters. 



Examinations of the occurrence and vitality of various weed seed species 



tinder different conditions, made at the Danish State Seed Testing Station 



during the years 1896-1923.* 



BY 



K. DORPH-PETERSEN, 



Director of the Danish State Seed Testing Station. 



The Danish journal " Tidsskrift for Landbrugets Planteavl " (Journal 

 •of Agricultural Plant Culture, vol. 17, 1910) contains a report, " Nogle 

 UnderB0gelser over Ulirudsfros Forekomst og Levedygtighed, udf0rte 

 -ved Statsanstalten DanskFrokontrol 1896-1910 " (Investigations of the 



Copies of this paper were circulated to the delegates in connection with. 

 Professor Bussard's lecture on weed seeds {see pp. 52-54). 



