5 
species, eight perfected more and four less than one-half of the 
good seed. 
Excluding Aster Nove-Anglie var. roseus and Solidago serotina 
(the poor seeds of which were only estimated), it was found that of 
the 17 remaining species, 1321 seeds were examined, 900 of which 
were good, and 471 or not quite 33 per cent. of the whole number 
were poor. The statements are all based upon a study of the 
seeds themselves before they were subjected to a test by germina- 
tion. An estimate of the quality of seeds by a study of the ex- 
ternal appearance of the seed itself must necessarily be uncertain, 
and is bound to overestimate the good seed. The test by germi- 
nation on the other hand may not give the full value to the vitality 
of the seed, and can never overestimate the good seed contained. 
In column 5 is tabulated the number of seeds germinated of the 
50 supposed to be good. Four of the 50 assorted or of the 80 
unassorted seeds of Aster Nove-Anghe germinated. 5 per cent. 
of unassorted, 8 per cent. of assorted, is the rate of germination of 
this plant. Of the seeds of /uula Helenium, 45 germinated. This 
gives as its rate 70 per cent. of unassorted and 90 per cent. of as- 
sorted. The percentages of these and the other species are tabu- 
lated in columns 6 and 7. 
Of the plants which may be considered as persistent field weeds, 
but two were grown. Their percentage of germination was very 
large, the first being the highest of all. 
EU LARD Oe here Oe en Ee JO 
Laraxacum officinale... . . umes oa 0 
The way side and fence row weeds in the ores of their percent- 
age of germination is as follows: 
TE 1 AE a ee oe a oO 
TS 0p SS PANS es ae se GD 
Licey Sin eR BO ae ae BG 
CONODSS Gewese eas f 58 
POITIER ee ee a BG 
Papa PAPCUADE es ee oe AD 
Aster Nove-ANghe i i ees 5 
Solidago serotina. . . . Pe 
The remainder, which may not i. Feaanteneed as weeds, range 
as follows: 
