AMPHIPODA OF THE SCOTTISH NATIONAL ANTARCTIC EXPEDITION. 459 
was little opportunity of distinguishing between characters subject to individual variation 
and those really common to the species. Consequently, when other specimens were 
obtained it was frequently found that they did not agree in all particulars with any of 
the species already described, and they were naturally considered to be new and were 
given a distinctive name. ‘This practice was perhaps the safest at the time, and it was 
the more desirable when the specimens came from a new locality ; but it unfortunately 
led to the idea that forms from fresh localities were almost necessarily new, and that 
the distribution of nearly all the species of Amphipoda was very limited. It also led 
to the introduction of long specific diagnoses, often containing characters of individual 
importance only. Naturally enough, specimens afterwards examined did not agree in all 
respects with these detailed descriptions, and thus a vicious circle was set up, leading to 
the continued establishment of new species, some of them being admittedly described in 
self-defence, and the fact that many species were widely distributed was long obscured. 
As knowledge gradually increased it was found that in many cases the same species 
had been described under various names, and the preparation of a general survey of 
the whole group, such as that for Das Tierreich, necessarily led to a considerable 
reduction of species. From the example of a few species which were readily recognised, 
and hence known to occur at places widely remote from one another, it was found that 
some species at any rate were more widely distributed than had been originally 
supposed. Much assistance in clearing up dittculties was obtained from the detailed 
study by various authors of individual species and the consequent elucidation of the 
various forms that occur in some species and especially of the differences between the 
sexes and of the changes that take place during growth; and it is to further work of this 
kind that we must look for assistance in defining the limits of the different species. 
Several of the species—or groups that I refer to under one specific name—are 
widely distributed in sub-Antarctic seas, and, as might be expected, the specimens from 
different localities now separate from one another are not always precisely the same, but 
show what may be considered local varieties. Some authors would doubtless prefer to 
eall these local varieties species and give each a distinctive name; but this must 
necessarily lead to an indefinite multiplication of species, with ever-increasing difficulty 
of determining those already established, and as a matter of practical convenience it 
seems to me to be better at present to endeavour to recognise these widely distributed 
species and to leave the determination of their varieties until a larger number of forms 
from many localities have been studied. 
In the list below I have indicated briefly the distribution of each species. From this 
it will be seen that an increasing number are now known to extend around the globe in 
sub-Antarctic seas, and that there is a greater resemblance between the Amphipodan 
faunas of South America, New Zealand, Australia, Kerguelen Island, and even South 
Africa, than appeared to be the case a few years ago. The importance of the facts 
on the question of the cause of this distribution cannot be discussed here. Another 
point made clear is that the number of species in northern seas represented by the same 
