BRITISH OYSTERS : OLD AND NEW. 219
Since the foregoing was set up, I have received from W. L.
Calderwood, Esq., F.R.S.E., of the Fishery Board for Scotland,
a number of examples of the older types of oyster once common
in West Scotland, and much valuable information. The speci-
mens sent me are from Jura and Loch Don in Mull, and are all
dead shells, and they clearly bring out that the form, which I have
called O. celtica, was the ancestral and predominant oyster of
these seas, the elongated outline gradually becoming rounder
as it passed southwards. The Jura shells are more disintegrated
internally than are those from Mull, which have the nacreous
lining preserved unbroken and lustrous. The existence of these
shells in Loch Don seems to have been unnoticed till discovered
by Mr. Calderwood, who found in the neighbourhood a series of
stumps of branches of trees roughly pointed at the ends. As Mr.
Calderwood found some of the shells attached to the stumps, they
may be, to use Mr. Calderwood's words, "regarded as collectors
to catch spat, as the Japanese use Bamboo" to-day.
Both in Jura and in Mull, other forms occur, but not so abun-
dantly as do the Celtica group. Amongst those from Mull is a
very fine example of the thick massive dark-coloured form of the
Estuarii group, like those mentioned above (p. 204), from
Grangemouth, Micklewood and the Nar Valley, which cannot
be differentiated from a Dogger Bank shell in my collection, all
being of the hippopus habit.
A few of the Mull shells both in size and ornament are not
unlike one of the shells figured by the authors of the Mollusques
Marines du Roussillon from Cancale, N.W. France (pl. i. fig. 4).
They are more recent than the shells I have referred to above.
The attempts hitherto made to re-populate the oyster beds
of W. Scotland do not appear to have been very successful,
whether in Shetland, or farther south, as at Arisaig in South
Invernesshire, where stock from Colchester, in Essex, was
laid down, or at Loch Sween, Argyleshire, where a large con-
signment from Arcachon, S. of Bordeaux, was utilised. Pro-
bably in these latter instances the climate was too strong for the
strains selected. The Loch Sween shell figured (pl. xvii., fig.
23), may have been introduced with the Arcachon shells.
The most recent of the Mull and Jura shells are of the
"Pandoure type" (plate xiii., fig. 6), and appear to be of no great
age from their condition, a portion of the adductor muscle still