18 
PACIFIC SCIENCE, Vol. IX, January, 1955 
(for example Tasmania) and is strong and 
prolonged— the extinction of the fauna will 
very probably occur. However, if the island 
is part of a long chain of islands, there is a 
good chance of reimmigration after the end 
of glaciation. 
3. If we have some large islands with their 
axis in a north-south direction, and these 
islands are isolated by very broad ocean areas 
from all continents and islands (e.g., New 
Zealand), the extinction or impoverishment 
of the fauna during strong glaciation is very 
probable. 
From these explanations, I think, it is evi- 
dent that the comparison of New Zealand 
with Australia or South America is not well 
grounded. The general conditions are very 
different. In New Zealand there is not the 
possibility of reimmigration of the fauna. 
Let us now consider the glaciation. In Aus- 
tralia there never was a glaciation which cov- 
ered extensive landmasses as by a shield. The 
glaciation touched only the mountain area 
and as a result no complete fauna or even the 
outstanding elements of parts of that fauna 
were destroyed. There was only temporary 
redistribution of the elements of the fauna 
in space. 
I think we have a similar condition also in 
New Zealand, where the glaciation was never 
so strong as to be able to destroy the whole 
fauna. Very possibly glaciation was the real 
cause of the extinction of the apiocerids, be- 
cause heavy snowfall usually is connected 
with diminishing temperature, and masses of 
snow and ice with low temperatures can cause 
the death of the apiocerids. 
But (and this is my main idea) it is not just 
the glaciation which destroys the whole fauna 
or large parts of it. Glaciation is the result of 
a specific combination of meteorological con- 
ditions when the snow accumulates more and 
more, beginning to form ice-shields and 
penetrating little by little into the valleys, but 
it does not mean the destruction of the fauna 
or the depression of its elements. 
In Patagonia and Chile we have at present 
the coexistence of a huge mass of ice and, 
at a distance of 2-3 hundred metres, a huge, 
very tall forest. Here we have equilibrium 
between the elements causing the glaciation 
and deglaciation, and this equilibrium has 
existed for at least 100-200 years because the 
forest near the masses of ice has required such 
time for growth. 
However, there is another type of meteoro- 
logical combination which can destroy the 
essential parts of the fauna without glaciation 
— the periodical or permanent lowering of the 
temperature in the winter. It is enough to have 
5-10-25 years of winters with the tempera- 
tures under 0° C, and all the insect pupae in 
the superficial layer of the soil will be frozen. 
The apiocerids are a very old group from very 
warm countries. They have not developed 
adaptations to cold winters, their pupae al- 
ways lie shallowly in the soil. A number of 
winters with temperatures under 0° C. can 
cause the complete extinction of the apiocer- 
ids. It will be useful to remember that in 
Siberia (Yakutsk area) in the coldest point of 
the whole world (the temperature in the 
winter drops to —70° C.) the snowfall is very 
small, usually so small that horses can find 
food under the snow during the whole winter. 
In such a way we can explain the extinction 
of apiocerids in New Zealand, not only di- 
rectly by strong glaciation but also by a series 
of very cold winters which can destroy those 
elements of the fauna without adaptation to 
low temperatures. In my review of Australian 
Apioceridae I mentioned only glaciation as 
the cause of impoverishment of the New 
Zealand fauna, but the above explanation 
should be taken into consideration, also. 
The presence of a very small number of 
species representing the three subfamilies of 
cyrtids in the New Zealand fauna shows us 
that the fauna is very impoverished. The main 
cause is probably the same as for the other 
families: glaciation, low temperatures in the 
winter, and high humidity. 
The remnants of the fauna show, however, 
rather high specialization. It is quite evident 
