244 MR. SPRUCE’S BOTANICAL EXCURSION 
white flowers, the petals barely equalling the sepals, and short pedicels 
reflexed in fruit; while in the one formerly sent (no. 375), the rose- 
coloured petals are twice the length of the sepals, and the long pedicels 
are eréct in fruit: there are also slight differences in the leaves. I 
have since found them growing together abundantly near the Indian 
village of Mayaca (or Mahica). You will have noticed that the anthers 
of Mayaca terminate in a tube one-third the length and diameter of the 
anther itself. In these two species the filament is short and tolerably 
stout. It has not been noticed that the calyx has a very short tube 
adherent to the ovary; its segments are truly valvate. The parts of 
the flower have assuredly the normal position: viz. the stamens oppo- 
site the sepals, and the carpels opposite the petals.—The habit of 
Mayaca is certainly different from that of the Commelynacee ; when 
out of flower its resemblance to Polytrichum juniperinum is most 
striking; yet its characters hardly suffice to sustain it in a separate 
Order. I have already sent you a true Commelynacea, with anthers 
opening at the apex. 
What I am about to relate will illustrate the utility of the large 
marsh grass, called Canna-rana (wild cane), of which I have already 
sent you specimens. As beef is only half the price at Santarem that 
it is at Para, a considerable traffic is kept up in cattle all the year 
round. A large portion come from the fazendas on the Furo de Sapucuá 
(near the mouth of the Trombétas), and the rest from the vicinity of 
Obidos, Almeirim, and Santarem. No supply of dry food is taken for 
their support during the voyage to Par’, but in leaving Santarem a 
quantity of Canna-rana is cut from the Ponta Negra, sufficient to last 
for two days, and all the way down a supply can be procured on the 
banks when required, until reaching the entrance to the channels below 
Gurupá, where there is no more Canna-rana, and the cattle subsist on 
the fronds of the Assai Palm until arriving at the city. The only 
objection to this latter food is that it causes the cattle to drink an 
inordinate quantity of water. So long as there is a sufficient supply 
. of the Canna-rana in the marshes near Santarem, the cattle thrive well ; 
but when they are reduced to eat other succulent grasses, they are con- 
stantly attacked with diarrhoea, and the cattle-keepers have not got into 
the way of giving them occasionally a little dry food. 
As I mentioned lately to Sir W. Hooker, Canna-rana is a chief con- 
stituent of the floating islands on the Amazon. Those we have seen 
