26
T. B . REYNOLDSON
these species to determine the climatic conditions at the time Britain
was separated from the continent. The evidence from lake-dwelling species supports O’Rourke.
The other fact which must be fitted into the hypothesis is the virtual
restriction of the western island faunas to P. nigra. Accidental introduction may be discounted; the arguments against have already been
stated, and to these can be added the uniformity itself, which suggests a
natural phenomenon. This means that either P . nigra survived glaciation in these small islands, extending as far north ;t9 Islay, an unlikely
event in view of the fact that its limit today is just a little farther north,
or that it dispersed more quickly than the others and reached the islands
before they were cut off from the mainland by the sea. There is no
obvious reason from structure or life-cycle why P. ?zigra should have
such an advantage, but if it is supposed that this species was mme widespread and abundant south of the glaciation than the others, just as it is
in Britain today especially in harsh habitats (Fig. 3, Table 11), then it
may have been able to follow the retreat of the ice more quickly. This
would be enhanced by its ability to occupy streams more readily than
the others (Reynoldson, 1956b); a view contradictory to that of Beauchamp (1932). Such an argument seems to be the only one which ean be
advanced to account for this characteristic distribution and provides
support for regarding the present-day pattern as a valid expression of
the ecological ranges of these species. The argument developed here presupposes that the western islands concerned (p. 21) were a11 linked to
the mainland at one stage after the retreat of the ice, and geological
evidence suggests that this is likely except for the Isle of Man; here
opinions are divergent (see Williams, 1962b). Such an explanation of the
distribution of P. nigra is supported by that of the stream-dwelling
Crenobia alpina and Polycelis felina. Both of these are more tolerant of
severe climate and poor habitats (Reynoldson, 1961c), and on the same
argument might be expected to follow the retreat of the ice even more
closely and quickly than P . nigra, so extending farther north. They may
even have been able to exist farther north than other species during
glaciation and Crenobia alpina has often been regarded as a glacial relict
species (Whitehead, 1922; Carpenter, 1928). Their present-day distribution supports such an argument for they occur not only on the western
islands supporting P. nigra but one or both also occur on islands slightly
farther north such as Mull, Rhum, Canna and Sanday (Bertram, 1939)
and on the main Orkney Island and Fair Isle. They have not been recorded from the more outlying islands such as Lewis (Elton, 1936;
R. C. Connolly, private communication), N. and S. Uist (R. C. Connolly, private communication) or the Shetlands (A. Milne, private communication). However, C . alpina was found on St. Kilda (J. B. Cragg-
T. B . REYNOLDSON
these species to determine the climatic conditions at the time Britain
was separated from the continent. The evidence from lake-dwelling species supports O’Rourke.
The other fact which must be fitted into the hypothesis is the virtual
restriction of the western island faunas to P. nigra. Accidental introduction may be discounted; the arguments against have already been
stated, and to these can be added the uniformity itself, which suggests a
natural phenomenon. This means that either P . nigra survived glaciation in these small islands, extending as far north ;t9 Islay, an unlikely
event in view of the fact that its limit today is just a little farther north,
or that it dispersed more quickly than the others and reached the islands
before they were cut off from the mainland by the sea. There is no
obvious reason from structure or life-cycle why P. ?zigra should have
such an advantage, but if it is supposed that this species was mme widespread and abundant south of the glaciation than the others, just as it is
in Britain today especially in harsh habitats (Fig. 3, Table 11), then it
may have been able to follow the retreat of the ice more quickly. This
would be enhanced by its ability to occupy streams more readily than
the others (Reynoldson, 1956b); a view contradictory to that of Beauchamp (1932). Such an argument seems to be the only one which ean be
advanced to account for this characteristic distribution and provides
support for regarding the present-day pattern as a valid expression of
the ecological ranges of these species. The argument developed here presupposes that the western islands concerned (p. 21) were a11 linked to
the mainland at one stage after the retreat of the ice, and geological
evidence suggests that this is likely except for the Isle of Man; here
opinions are divergent (see Williams, 1962b). Such an explanation of the
distribution of P. nigra is supported by that of the stream-dwelling
Crenobia alpina and Polycelis felina. Both of these are more tolerant of
severe climate and poor habitats (Reynoldson, 1961c), and on the same
argument might be expected to follow the retreat of the ice even more
closely and quickly than P . nigra, so extending farther north. They may
even have been able to exist farther north than other species during
glaciation and Crenobia alpina has often been regarded as a glacial relict
species (Whitehead, 1922; Carpenter, 1928). Their present-day distribution supports such an argument for they occur not only on the western
islands supporting P. nigra but one or both also occur on islands slightly
farther north such as Mull, Rhum, Canna and Sanday (Bertram, 1939)
and on the main Orkney Island and Fair Isle. They have not been recorded from the more outlying islands such as Lewis (Elton, 1936;
R. C. Connolly, private communication), N. and S. Uist (R. C. Connolly, private communication) or the Shetlands (A. Milne, private communication). However, C . alpina was found on St. Kilda (J. B. Cragg-