New York AGRICULTURAL EXPERIMENT STATION. 185 
sown is low in viability or else the mortality of the seedlings is ex- 
tremely high. 
IMPURE AND ADULTERATED SEED. 
New York farmers experience considerable difficulty in obtain- 
ing pure alfalfa seed. It being next to impossible to produce al- 
falfa seed in New York, all of the seed used must be purchased 
from seed dealers. Frequently, the Station is asked to recom- 
mend some seedsmen who may be relied upon to furnish good seed. 
Such requests are refused because it is the established policy of the 
Station to avoid advertising anyone’s business. Farmers are ad- 
vised to buy by sample and not to sow alfalfa seed until an expert 
has examined it for dodder and other troublesome weeds and for 
evidence of adulteration. For residents of New York, the Station 
makes such examinations free of charge. 
IMPURITIES. 
Between November 20, 1906, and August 15, 1908, the Station 
analyzed 548 samples of alfalfa seed sent in by farmers and seeds- 
men from various parts of the State. Since these samples were 
grown in many different parts of the world it is to be expected that 
they would contain many different kinds of weed seeds and such 
was found to be the case. Several of the weed seeds we were un- 
able to identity. One of these unknowns, found in 37 samples, was 
sent to the Seed Laboratory of the United States Department of 
Agriculture for identification. Mr. Edgar Brown, Botanist in 
Charge, informs us that the seeds are of a species of Centaurea, 
which is common in Asia Minor, and that they probably come to 
us only in the seed of Turkestan alfalfa. Thus it is sometimes 
possible to determine where seed was grown by the weed seeds it 
contains. 
Usually, it is the character of the impurities rather than their 
quantity which makes them objectionable. The most objectionable 
impurity of all is dodder seed. So little as one dodder seed per 
pound renders alfalfa seed unsafe for use. Dodder often makes 
serious trouble in alfalfa fields and no chances should be taken with 
it. Of the 548 samples analyzed, 126, or 23 per ct., contained more 
or less dodder. In 122 of the infested samples the number of dod- 
