40 95 Sigurd Johnsen. 
as this is an almost closed-in sea it took some time before the 
new conditions were established. Moreover during the melting 
away of the ice the amount of freshwater added by rivers must 
have been greater than at the present time. The exchange of 
water through the Straits of Gibraltar must have been quite the 
opposite one to that of the present time viz. a surface-current 
of low salinity running out and an under-current of Atlantic 
water coming in. Thus the temperature and the salinity of the 
Mediterranean waters were gradually raised to the values of the 
Atlantic waters. During this period we must suppose that the 
race thori was established in its main features; the currents in 
the Straits of Gibraltar, as outlined above, were not favourable 
to its re-entering the Atlantic. Owing to the decreasing amount 
of freshwater from rivers and rainfall and to the extensive eva- 
poration, tke salinity was in the following period gradually 
raised above the values of the Atlantic water; the currents in 
the Straits of Gibraltar changed direction and the hydrographical 
conditions become as they are known at the present time. Once 
during this period the separation in a western and an eastern 
stock of M. glaciale thori took place, the species retreating to 
the areas with the lowest values of temperature and salinity. 
It does not seem that the Mediterranean race during this second 
period has been able to regain ground in the Atlantic. Certainly 
some specimens probably belonging to the Mediterranean race 
have been captured along the Atlantic coast of the Peninsula 
(cir. p. 21), but as larval stages and young-fish are not known 
in these waters the specimens in question may be regarded as 
occasional migrants from the Mediterranean; possibly also some 
young-fish may be carried out in the Atlantic by the outflowing 
under-current. 
The hypothesis here set forth, that M. glaciale in the Medi- 
terranean is a relict form might be tested by a more comprehensive 
comparison between the pelagic fish-fauna of the Mediterranean 
and the Atlantic. At present this cannot be satisfactorily done. 
Our knowledge of the Mediterranean fish has been much added © — 
to by the Danish investigations in the , Thor“ of late years. On 
the expeditions in the Atlantic of the various Danish research- 
vessels and of the “M. Sars” and the “Arm. Hansen” the 
investigations have been carried out with fishing-gear, especially 
