78 
or pericarp. The spikelets become easily detached when ripe, 
drop into the water and leave the bare spindles standing up stiff 
like spears until they break down along with the stems, which 
gradually decay during the winter and spring. The ripening of 
the grains takes place mostly in October. Abundant as the 
flowering is year after year, the fruiting is very uncertain. In 
October, 1907, I found only few perfect grains of Townsend’s 
grass in the Isle of Wight, whilst after the hot summer of 1911 
there must have been millions of them in the Lymington beds 
alone. In front of the Pylewell Estate and to the east of it, there 
was, at high water-mark, a long band of chaff, which, on 
examination, consisted almost entirely of spikelets of the grass. 
A good handful taken up at random numbered 1882 spikelets, 
of which, however, only 193, or a little over 10 per cent, con¬ 
tained apparently sound grains. They were spikelets cast out. 
Most of those which had good grains must have sunk on the 
mudflats. Samples taken up from 1-1A feet below highwater-mark 
showed a percentage of over 60 good grains in the 100. 1912 
was again a bad year for the maturing of the grains. The bottom 
of the mudflats is overrun with a delicate film of*filamentous Algae, 
and it is by their fine threads that the spikelets when sinking are 
caught and gradually so fixed that they remain there until they 
germinate, when the seedling becomes quickly and firmly estab¬ 
lished by the rapid growth of the roots. 
I he presence of mud is an important factor in the forma¬ 
tion of colonies of Townsend’s grass. It need not be necessarily 
pure mud, although the grass is most generally and abundantly 
found on such, but if there is an admixture of shingle and sand 
it must not exceed a certain amount. As all the mudflats are 
formed in comparatively quiet waters, we find the grass confined 
to these and entire!}' outside the zone of heavy surf and strong- 
tides. On the flats themselves the limit to which the grass can 
spread is determined by the depth at high tide. Although the 
grass can stand much submersion, z \-3 feet of water at ordinary 
high tide seem to be the maximum it can bear. This 2^-^ feet 
line determines the outer limit suitable for the grass. The inner 
or landward limit lies close to, but still clearly below highwater- 
mark. This is also true of backwaters and the estuaries of rivers. 
Within those limits the grass has a practical monopoly in so far 
as flowering plants are concerned, no other species associating 
with it, except in places Spartina alterniflora. 
The area occupied by the grass on the South Coast of Eng¬ 
land extends at present from the entrance to Chichester Harbour, 
near \\ est \\ ittering in the East to the Southern and Western 
shores of Poole Harbour in the West, covering several thousand 
acres, the largest beds being those on the West side of Southamp¬ 
ton Water, in Lymington Harbour, in Hurst Castle Bay, and at 
the mouth of Beaulieu River. The Spartina bed between Hythe 
pno Cal shot Castle, probably the oldest and the most compact, 
is about fi\ e miles long and j.00-600 vards wide, and covers over 
