Pelagic Hyperiidea — Bary 
327 
per haul over 100 fathoms (182.9 m.) of water 
increases slightly. Captures of both P. aus- 
tralis and P. gracilipes increase to February and 
decrease to March. The T-S-P diagram dem- 
onstrates, however, that the two species are 
resident in coastal water, and it is probable 
therefore that the February-March (seasonal) 
decrease is not more than a contributing fac- 
tor to the lower number of specimens cap- 
tured from the deeper water. Thus, for the 
most part there appears to be a considerable 
degree of independence between the num- 
bers of a species captured per haul per month 
and the numbers captured relative to depth 
of water. A similar condition is also suggested 
by the averages of all the common amphipods 
(Fig. 7). The average catch increases to 100 
fathoms in a similar manner to that for Jan- 
uary and February. However, the catch from 
water deeper than 100 fathoms continues to 
increase, although at a lesser rate, while the 
catch for March drops sharply. 
DIURNAL VARIATION 
Vertical migration undoubtedly occurs 
among species of Amphipoda (e.g., Stephen- 
sen, 1925; Hardy and Gunther, 1935; Bous- 
field, 1951). However, other factors than light 
intensity would appear to be operative in the 
control of vertical distribution. Bousfield sug- 
gests quality and quantity of food, age of 
specimens and predation, and implies that the 
results of vertical migration may be modified 
by these. There may be vertical migration in 
connection with propagation as well (Steph- 
ensen). Stephensen also discusses the point 
that a number of species occur singly, or in 
groups of a few specimens. He considers that 
this applies to almost all deep-sea species, and 
also to some surface species. As well, there 
appears to be an inherent tendency for some 
species to shoal (Hardy and Gunther, 1935). 
These several factors, some of which ap- 
pear to be interrelated, almost inevitably 
would result in anomalous occurrences in 
terms of a normal diurnal rhythm. With spe- 
cies tending to appear somewhat irregularly 
at the surface, patchiness in their occurrences 
would result. Hardy and Gunther (1935) dis- 
cuss such patchiness among amphipod spe- 
cies and the effects this has on their distribu- 
tion patterns, with particular reference to the 
area about South Georgia. 
The data on the species in the present 
study demonstrate that vertical migration 
takes place and that patchiness is frequently 
met. The former is reflected in diurnal varia- 
tion of numbers at the surface, and the latter 
in the great variation in the numbers captured 
from hour to hour (Figs. 8, 9), or station to 
station, of all the common species (Figs. 5, 
6) . The average haul for each hour is derived 
from the average of one to five stations. In 
contrast the average number of tows per hour 
is low (almost three per hour), which pre- 
cludes the present data from being fully rep- 
resentative of the patterns of diurnal variation 
in New Zealand waters. 
However, the numbers of the five com- 
monly captured species, and of the juveniles 
of Parathemisto spp., increased at night. Para- 
themisto gaudkhaudii and juveniles of Parathe- 
misto spp. exhibited pre- and post-midnight 
increases in numbers; although such is more 
or less typical of the diurnal behaviour of zoo- 
plankton (Cushing, 1951), the post-midnight 
rise occurs between 0100-0200 hours, which 
in southern New Zealand is several hours too 
early to be considered a predawn rise. The 
other species increase to a single peak during 
darkness either prior to {Hyperoche mediter- 
ranea), or subsequent to midnight {Cyllopus 
magellanicus, Parathemisto australis^ P. graci- 
lipes and Parathemisto juv.). A large catch of 
several species, made at Station 189 at 0125 
hours, is responsible for the peak of numbers 
taken between 0100 and 0200 hours for P. 
gaudkhaudii, P. australis, P. gracilipes, and 
Parathemisto juveniles. A total of four stations 
were occupied between 0100 and 0200 hours, 
but only at Station 189 were any of these spe- 
cies captured. This may be a result of patchi- 
ness in distribution. On the other hand. 
Station 189 was in mixed coastal-subantarctic 
