GEOGRAPHIC DISTRIBUTION OF THE PACIFIC SALMON 683 
planted in a Canterbury stream, yet the Canterbury rivers now provide the best quinnat fishing in 
the Dominion, the species having migrated to their mouths from the Waitaki, where the original 
fry were planted. The Wairau has been fairly generously stocked and yet shows no appreciable 
run of fish. The inference is that it is probably too far north — outside the influence of the purely 
Antarctic water which attracts the bulk of the species — though an odd few are known to run into 
the Wairau, and, in fact, into some of the southern rivers of the North Island. This season an 
indubitable quinnat was caught in the Tukituki River, Hawke’s Bay. It does not follow that these 
parts are suitable for the permanent establishment of the species in abundance. The limit to which 
the influence of hydrographical factors pertaining to Antarctic waters extends will doubtless vary 
at different times, and it may be that in odd years the Cook Strait neighborhood, or even farther 
north, may provide suitable and congenial conditions for the quinnat salmon. But the indications 
afforded both by experience and by theoretical considerations seem to emphasize the probability 
of the fundamental relationship between the nature of the sea-water and the distribution of these 
salmon. There is also the case of the attempted acclimatization of the quinnat in the Hokitika 
River, on the west coast of the South Island. Our departmental reports show that between 1910 
and 1924 the fry from over three million ova were planted in the head-waters of this river. The 
only apparent outcome has been a stock of lake-dwelling quinnat which has established itself in 
Lake Kanieri. As is well known, the west coast of the Dominion is washed by a warm current 
which has eddied across the Tasman Sea from the coast of eastern Australia, and which was origin- 
ally a branch of a westerly-trending subequatorial current. Again it seems to be a case of the 
wrong sort of sea-water for a salmon species. Where the quinnat smolts, which have presumably 
entered the Tasman Sea to the number of thousands or hundreds of thousands, have disappeared 
to is a mystery which may never be solved. This discussion is admittedly somewhat speculative, 
but it seems necessary to ventilate these considerations in view of the frequent recommendations, 
based rather on what is desirable than on what is probably feasible, to stock this or that river with 
salmon. 
The analysis of the physical and chemical properties of the coastal waters of 
New Zealand bears out Mr. Hefford’s assumptions as to the unfavorable character 
of the marine waters off the north and west coasts of South Island and the entire 
coast of North Island. The directional drifts of the ocean currents, the mean seasonal 
surface temperatures, the mean annual subsurface temperatures at 200 meters, and 
the mean annual surface salinities are given in figures 1-4, respectively. These data 
were calculated in the same manner as those for the North Pacific and were taken 
from the studies of Buchan (1894) and Schott (1928 and 1935). Schott (1935) gives 
a complete summary of all the hydrographic data collected in this region. 
The directional drifts of the ocean currents shown in figures 1 and 5 were cal- 
culated for the months of January, February, and March, which cover the spawning 
migration period of the Pacific salmon in this region. The South Pacific Current and 
the Antarctic Drift, which are so closely related that they may be considered as one 
current, carry waters of low temperatures and salinity. South of New Zealand a 
portion of this combined current divides. Part flows northward along the west coast 
of South Island and merges with a branch of the warm East Australian Current near 
the central coast of the island. Natural runs of California chinooks have been estab- 
lished only in the streams along this coast south of the point of confluence of these 
currents. The remainder flows northward along the east shores of South Island to 
Cook Strait where it is met by counter drifts from the warm South Equatorial Cur- 
rent. Natural runs of California chinooks have also been established only in the 
streams along this coast south of the point of confluence of the cold and warm currents. 
This shows a relationship between ocean drifts and the occurrence of natural runs of 
salmon similar to that found along the coasts of Korea and Japan. 
The mean surface and subsurface temperatures, and mean surface salinities, 
given in figures 2, 3, and 4, all show that the coastal waters of North Island are 
warmer and more saline than those of South Island. The mean 15° C. surface and 
10° C. subsurface isotherms and mean 35 parts per thousand surface isohaline all 
touch the shores of South Island near the upper limits of the range in which the sal- 
