I70 
F. TRACY HUBBARD 
In testing the stability of the various characters I found that 
ordinarily the second glume of 5. viridis equaled the sterile lemma in 
length whereas in 5. italica it was for the most part noticeably shorter. 
Furthermore the spikelets in 6*. viridis, as soon as they become at all 
mature, shell out the whole spikelet — quite readily — leaving a cup- 
like receptacle; whereas those of S. italica shell out the fruit^ only 
leaving the first and second glume and sterile lemma behind. This 
appears to me to be a good specific character and the use of it as a key 
character in separating the specimens gave homogeneous groups. 
The name germanica was first used by C. Bauhin, Theatr. Bot. 518, 
fig. (1658) as Panicum Germanicum sive panicula minor e. Exactly 
what form of S. italica Bauhin had is impossible to state though he 
undoubtedly had one of the larger lobulate forms as he states that the 
panicle is nine inches rarely a foot long. It may be yellow, purple 
or black, but he fails to say anything about the length of the setae. 
The plate is rather indefinite, but represents panicles with short setae. 
While the name originated with Bauhin the first use of it after 1753 
is Panicum germaniciim Miller, Gard. Diet., ed. 8, no. i (1768). 
Miller describes the species as follows ''Spica simplici cernua, setis 
brevioribus, pedunculo hirsuto," basing his name on C. Bauhin, 
Pinax 27 (1671). In the discussion of his species of Panicum he 
makes the following statements regarding P. germanicum, "The stalks 
are terminated by compact spikes, which are about the thickness of a 
man's finger at their base, growing taper toward their points, and are 
eight or nine inches long,^ . . . ." The phrase ''setis brevioribus" 
has undoubtedly led to the recent interpretation of S. italica var. 
germanica as a form with setae shorter than the spikelets, but does this 
phrase compare the length of the setae with the spikelets; why may it 
not be a term comparative with the following species, Panicum 
italicum, which according to Bauhin's plate in Theatr. Bot. 519 is a 
plant with long setae? I believe that this interpretation is correct as 
I have a photograph and spikelets of the Miller material in the British 
Museum which were kindly sent me by Mr. A. B. Rendle. This 
material was collected in Chelsea Garden in 1760 and is labeled 
"Panicum Germanicum sive panicula minore" and consists of a 
panicle and two leaves. As a centimeter scale has been placed on the 
^ The term fruit is consistently used in this paper to mean the seed inclosed in 
the fertile lemma and palea. 
- These measurements are too long for subvar. germanica and also do not agree 
with Miller's specimen. 
