124 
great number of umbels matured, and the constricted access of 
light, suggested that there might have been an undue strain upon 
its potent energy, and that the fruit produced would be found 
greatly defective and much diminished. For this reason, and as a 
matter of general interest, a comparison was made with specimens 
growing in the wild state, in open ground, under the best natural 
circumstances, and producing the finest umbels. Umbels of the 
axis were taken from the best specimens found in the neighbour- 
hood, and from the variety marztima on the Lizard Peninsula, 
Cornwall, where the plant was abundant and vigorous. 
It will not be necessary to go into the details of this analysis 
beyond giving an account of the method pursued, and mentioning 
some of the general results. 
A, Garden grown specimen. 
In the general method adopted in examining this and the 
other specimens, the surface of the umbel was noticed as in 
some degree indicating the degree of ripeness of the seed. 
The number and condition of the bracts were observed as 
apparently bearing some relation to the productive potency of 
the umbel or umbellules. The umbellules were separated 
into series from without inwards, as nearly as could be, in 
successive rings, and the seed of each was separately divided 
into good and bad, the fully developed mericarp was taken as 
good, the imperfect, shrivelled, or injured, as bad. When the 
two mericarps advanced very little beyond the floral state of the 
ovary, and remained united and inseparable, they are reckoned 
one bad seed. 
Resvtts oF ANALYsIS oF A. 
There were 12 ridges in the stem of the umbel: 11 
terminated in pedicles or rays of the umbellules, with bracts 
at the base, 1 ended in a ray without a bract, and was placed 
nearer the 2nd than the 1st ring; 38 of these bracts were 
bipinnate and 8 pinnate. The umbel was concave on the fruit 
surface, the outer row of umbellules had the surface turned 
inwards, and somewhat downwards, towards the centre. 
The umbel was composed of 71 umbellules, 
which produced a total of ...... 3,614 seeds. 
Of these there were good ....,. 1,911 ,, 
‘ Dyadetas tks. L703" *,, 
Thus the good exceeded the bad by 208. 
The relative proportion was 26°91 to 23°98. 
In the outer rings of the umbellules more good seed was 
eu than bad; in two, however, the bad exceeded the 
good. 
