IMPROVEMENT OF THE SUGAR CANE. 107 
lar error arose it is difficult to tell. Some years ago, 
with the assistance of Mr. CARRUTHERS, the Keeper of 
Botany at the British Museum, I searched through a lot 
of old botanical works and we could find no trace of this 
belief, until about 1750 when HUGHES wrote his Natural 
History of Barbados. Nay, more, we found descriptions 
of the seed of the sugar cane so complete and accurate 
that CARRUTHERS considered they must take precedence 
of all recent work. Dr. DUTKONE in 1790 described the 
floresence of the cane as follows :— 
“It’s characters are these: it has no empalement, but 
a woolly down, longer than the flower that encloses it; 
the flower is bivalve; the valves are oblong, acute- 
pointed, concave, and chaffy; it has three hairlike 
stamina, the length of the valves, terminated by oblong 
summits, and an awl-shaped germen, supporting two 
rough styles, crowned by single stigmas; the germen 
becomes an oblong, acute-pointed seed, invested by the 
valves.’ In figure 27 of the plate in PORTER’S work on 
the sugar cane the germen with its two styles and stig- 
mas is shown. 
In the report of the Barbados Botanic Station for 
1889 I gave the local history of the knowledge of the 
produétion of fertile seeds of the sugar cane, and in 
a letter to the Manchester Examiner in September, 
1890, I gave the history of the matter as far as I 
knew it, This letter was re-published in the Kew 
Bulletin and afterwards copied in “Sugar” for Feb- 
ruary, 1891, Since then I have come across a few 
instances showing that certain investigators in the 
Western tropics were aware of the occasional fertility 
of the seed of the sugar cane, for instance Baron 
02 
