586 
Journal of Agricultural Research 
Vol. XXXI, No. 6 
leaves on both was about the same. The difference between them 
was evidently due to the greater internode length of the former 
(pi. 1). It was at once suggested that this difference might be util¬ 
ized as a means of identifying the seed of the annual then coming 
into prominence and the seed of which is indistinguishable from that 
of the biennial. 
On October 11, when 60 days old, these seedlings were taken up 
and measured from: (1) Cotyledons to tip of the longest leaf, and 
(2) cotyledons to growing point of the stem. Naturally, the exact 
location of the growing point could not be ascertained without 
elaborate dissection in each case. However, by careful observation, 
its position could be determined very closely, and the writer is con¬ 
fident that the error Will not exceed ±0.5 mm. 
In the seedlings of Melilotus, as in Trifolium, Medicago, and many 
other legumes, the first leaf is unifoliolate (pi. 1, a) and all subsequent 
leaves trifoliolate. Counting the unifoliolate leaf as one, the number 
of expanded or nearly expanded leaves is equal to the number of inter¬ 
nodes above the cotyledons, though the upper internode may not 
yet have reached full length. By noting the number of expanded or 
nearly expanded trifoliolate leaves, a means is afforded for comparing 
seedlings of equal development. 
In the work here reported, all seedlings taken up on a given date 
were separated into groups, each group containing thosie plants with 
the same number of expanded, or nearly expanded, trifoliolate leaves. 
The last internode lengthens as the upper leaf expands, and a part 
of the irregularities in the measurements recorded in Table I is due 
to unequal maturity of the upper internodes. The cotyledons 6 re¬ 
main attached for many days or weeks and afford a convenient and 
fixed joint from which to measure (pi 1, 6). The measurement from 
cotyledons to tip of longest leaf was found to be unsatisfactory, 
because of great variations in petiole lengths, and this measurement 
was abandoned. In measurements made subsequent to those re¬ 
corded in Table I the length from cotyledons to first (unifoliolate). leaf, 
and the length from cotyledons to growing point were made. In all, 
4,610 seedlings were measured, involving more than 9,000 meas¬ 
urements. 7 
Table I .—Average stem and internode lengths in millimeters of seedlings measured 
October 11, 1921, 60 days after date of seeding 
Annual 
Biennial 
3-leaved ° 
4-leaved 
5-leaved 
3-leaved 
4-leaved 
5-leaved 
Average stem length____ 
22.4 
5.6 
35.8 
5.16 
45.55 
7.6 
7 
1.75 
10.8 
2.16 
11.1 
1.85 
Average internode length _ 
« The number of internodes was one more than the number of trifoliolate leaves. 
It will be seen from Table I that the length of the internodes in the 
annual was two, three, or more times as great as in the biennial (pi. 1). 
These facts were reported at a meeting of the Association of Official 
Seed Analysts of North America, held in July, 1923. 8 
6 In all plates showing mounted specimens, the plants have been so mounted that the cotyledons, or 
the point where they were attached, are on the heavy black line. 
7 The writer’s thanks are due to L. W. Kephart for assistance in making these measurements. 
* Association of Official Seed Analysts, proceedings of the fourteenth and fifteenth 
ANNUAL MEETINGS. 148 p. 1923 
