336 Annual Report of the Consulting Botanist for 1883. 
the standard of ninety per cent., but only a quarter of the samples 
of white clover were up to the standard, though none fell below 
sixty-four per cent. 
A good-looking sample of seed-wheat purchased by a 
member, which failed to come up, was found on trial to be 
quite dead. The seeds were perfect and well filled, but the 
embryo had lost its vitality in every case, from the grain having 
been kept too long. It cannot be sufficiently impressed on 
sellers and buyers of seeds that wheat usually loses its vitality 
when kept from six to ten years ; and that under special cir- 
cumstances it may become useless even before it is six years 
old. No real gain is secured by keeping seed-wheat over the 
first sowing-time after it has been reaped. 
It has been objected by some members of the seed-trade that 
the germination in the laboratory is a very unsatisfactory test 
of the ffrowingf-value of the seed. It is asserted that much better 
results are to be had from the sowing under natural conditions 
in the field than can be obtained by the artificial conditions 
under which germination is usually tested. These statements 
are very plausible, but they are utterly wrong. The conditions 
necessary to the germination of a seed are all secured in a 
proper testing-case, and none of the accidents which cause the 
death of the seed or of the young plant are permitted. Too 
deep immersion in the soil, too little water, too much sun, the 
attacks of insects, and other conditions met with in the field, or 
even in an experimental plot, are avoided in the laboratory. 
But the actual experiments of Dr. Sturtevant, the Director of 
the New York Agricultural Experimental Station, have esta- 
blished beyond controversy that the proportion of seeds from 
the same sample which vegetate in the garden are considerably 
smaller than those that manifest vitality in a germinating appa- 
ratus. The test of germination is really a test of the seed's 
vitality. The first evidence of this is the sending out of the 
rootlet. And the appearance of the radicle or rootlet is suffi- 
cient to determine the capacity of the seed to germinate. The 
tender plant may have a short supply of food in the seed, or it 
may have difficulty in getting such a hold of the soil as to 
obtain food for itself, or many things may happen to arrest its 
progress and cause early death. Dr. Sturtevant found no con- 
stant relation between the percentage of seeds which germinated 
in the laboratory and those which produced young plants in the 
garden. In some exceptional cases, seeds, every grain of which 
germinated in the laboratory, were unable to produce a single 
plant in the field. Such cases were met with only in respect of 
some seeds which were very long in vegetating, and were con- 
sequently subject to more accidents, and the attacks of more 
