GEOGRAPHIC DISTRIBUTION OF THE PACIFIC SALMON 
681 
sea -run populations and even here only within certain limits. The streams on both 
islands are shown in figure 5. The streams on South Island which support natural 
runs of chinooks are indicated by a solid circle, and those which have been stocked 
frequently from 1910 to 1929, but which have never developed natural runs, are 
indicated by a solid triangle. Does the explanation of this failure of the chinooks 
to develop natural runs in certain streams on South Island and in none of the streams 
on North Island lie in unfavorable conditions in the fresh- water or the marine 
environments? 
The streams on both North Island and South Island are quite similar in origin 
and type. Most of the larger ones originate in mountain lakes and flow rapidly to 
the sea over gravelly and rocky beds, see figures 6-8. Percival (1932), in describing 
the streams of New Zealand, states: 
The geological youthfulness of the present land-surface of New Zealand accounts for the rela- 
tive absence of slowly flowing rivers such as, in other countries, give shelter to a great variety of 
free-swimming organisms and allow of the growth of much vegetation on the bed. 
dfe H: ije 5|e sH # 
Generally speaking the rivers of New Zealand are comparable with the portions of the Euro- 
pean rivers called by Thienemann (28) “Aschenbach” (Grayling stream), where the bed is stony 
and liable to flooding through the accumulation of surface water. 
The streams on North Island, owing to the milder climate, are somewhat warmer 
than those on South Island. At Rotorua, North Island, the mean air temperature 
for January is 18° C. and for July is 7.5° C., while at Queenstown, South Island, the 
mean air temperature for January is 15.5° and for July is 3° C. Phillips (1929) 
reports stream temperatures on North Island, during the winter and spring, as low 
as 8° C., and during the fall from 12.5° to 15.5° C. Hobbs (1937) reports the mean 
monthly temperatures of salmon-bearing streams on South Island as ranging from 3° 
C. in midwinter to 16.5° C. in midsummer. Percival (1932), in discussing the presence 
of fish food in the streams on both islands, states that it is sufficiently abundant in 
most of the streams to support trout and other fresh-water fishes. In view of these 
facts it may be assumed that the streams on both islands provide favorable environ- 
mental conditions for the survival of the salmon during their fresh-water existence. 
An examination of the environmental conditions found in the coastal waters of 
the islands, however, gives an altogether different picture, for North Island is almost 
wholly bathed by a tropical current and South Island by an Antarctic current, The 
directional drifts of these currents are shown in figures 1 and 5. 
Hefford (1929), in discussing the reasons why runs of chinooks have not been 
established in the Wairau and Hokitika Rivers on South Island, and in all of the 
rivers on North Island, makes the following statement: 
It is known that off the south-eastern coasts of South Island the water in the sea is of Antarctic 
origin. There is a general set or drift in a north-easterly direction of cold water from the south, 
and this water produces the prevailing conditions in the sea off the Otago and Canterbury coasts 
where the quinnat have been established for some years. The South Equatorial Drift, which sets 
from the eastward and impinges upon the east coast of North Island, may be said to dominate the 
conditions to the northward of East Cape; while between that point and Cook Strait there is a mix- 
ture of this subtropical water with water from the south. For a long time navigators have been 
familiar with these “sets” or surface movements of the sea, but it was not until the hydrographer 
of the Danish research steamer Dana had applied physical and chemical tests to the water sampled 
at intervals between the east coast of Auckland and the coast of Otago, in January 1929, that the 
significant differences in the character of the water along this line were ascertained. It seems clear 
from the Dana's observations that the present distribution of quinnat salmon off the New Zealand 
coasts coincides with the occurrence of practically unmixed Antarctic water, with its characteristic 
physical and chemical qualities. Not a single individual of the quinnat species has ever been 
