192 REPORT OF THE DEPARTMENT OF BOTANY OF THE 
ants of alfalfa seed,” but for farmers this is scarcely worth while. 
Most farmers will find it safer and more satisfactory to rely on 
tests made by experts at the Experiment Station or in the United 
states Department of Agriculture. An analysis made by an expert 
is the farmer’s chief safeguard against impure and adulterated seed. 
Samples of alfalfa seed for analysis should contain at least two 
ounces and should be taken in such manner as to represent fairly 
the lot of seed to be tested.*° If the seed is in bags the sample 
should contain some seed from each bag and in all cases a portion 
of it should be taken from the bottoms of the bags. When alfalfa 
seed containing dodder seeds is shipped in bags in an upright posi- 
tion the dodder seeds, being smaller than the alfalfa seeds, tend to 
settle to the bottoms of the bags. Samples should bear the name 
and address of the sender. The postal rate on seeds is one cent gn 
ounce. 
Many of the samples received at the Station are entirely too 
small for a dependable analysis as regards dodder. The samples 
sent out by seedsmen are usually too small. A majority of the 
548 samples reported upon in this bulletin contained less than one 
ounce. Had all the samples been of proper size it is likely that the 
number found to be infested with dodder would have been larger. 
A sample from Moira, N. Y., containing only 2 grams was free 
from dodder. Upon being informed that the sample was too small, 
the sender forwarded aiother 20-gram sample of the same seed. 
This contained 22 dodder seeds which is at the rate of 499 per 
pound. | 
A Canandaigua farmer brought to the Station a good-sized sam- 
ple taken from the top of a bag of seed. It was entirely free from 
dodder. A few days later he brought another sample from the 
same bag but obtained by thrusting his hand deeply into the bag. 
This sample contained dodder at the rate of 20 seeds per pound. 
A farmer at Delhi, N. Y., sent a sample (5.3 grams) which was 
found to be free from dodder. Upon receipt of the Station report 
on the sample he purchased and sowed some of the lot of seed from 
which, supposedly, the sample had been taken. However, the seed 
dealer was in doubt as to which bag the sample had been taken 

- ™Some bulletins especially useful for this purpose are the following: Hill- 
man (44), (45), (46); Brown (9); Roberts and Freeman (83). 
* Detailed directions for taking samples are given in, Rules and Apparatus 
for Seed Testing. U. S. Depart. Agr. Office of Experiment Stations. Cire. 
34 (Revised). 1906. 
