186
THE ESSEX NATURALIST.
detached specimens they were practically always seen to have
minute particles of extraneous matter adhering to them (fig. 4).
That the adhesive material was indeed quite a strong cement was
further proved by rolling the little clusters of sand-grains about
under the cover-glass when it was found almost impossible to
break up the clusters, and that in any rearrangement of the
sand-grains relatively to one another the gaps between them
were still bridged by the algal cells. But although obviously
provided with a very efficient adhesive, it was scarcely to be
expected that such small bodies as the cells of S. antennatus,
which are at most only about 1/1500th of an inch in length, could
possibly hold together the larger sand-grains, and this indeed
was found to be the case. No particles of sand much over
1/100th of an inch were ever seen in the flocculent masses. The
average size of the grains may perhaps be taken at about 1/400th
of an inch, but the majority were smaller than this, ranging
down to the merest specks even under high magnification.
The phenomenon above referred to continued to be more or
less in evidence from the time when it was first noticed, namely,
towards the end of June, 1927, until the middle of November.
It appeared again in the middle of March, 1928, and was observed
at intervals during the year until the end of October. By
mid-November no more of the flocculent masses of sand-grains
could be found. In every case when the material was examined
it was found to be as described above, the only alga present being
S. antennatus.
Another instance of sand-grains agglutinated by small algae
has been observed on the sandy part of the margin of a pond by
the side of Earl's Path, leading from Loughton to High Beach.
In this case, however, the little flocculent masses were not very
abundant, and they could scarcely be said to produce a "field"
phenomenon. Nevertheless they were essentially the same as
those occurring in the Hollow Pond, except that the agglutinating
agents comprised several different species of algae. The most
abundant form in this case was Pediastrum glanduliferum Bennett
but some Scenedesmus antennatus were also present and a few
other small unicellular forms which were not definitely identified.
It is presumably no mere coincidence that the processes on the
marginal cells of P. glanduliferum are expanded into little knobs