395 
Mosquito Larvae of New Guinea — BiCK 
collector’s initials. Ecological terms were 
listed around the margins so the collector 
simply checked the appropriate blocks. This 
method reduced the possibility of omitting 
important data from field notes compiled at 
random. A similar card punch system is dis- 
cussed by Boyd (1930). 
Identifications 
For most collections, a representative series 
of individuals was reared in isolation. Identi- 
fications in nearly every case were based on 
reared material. The majority of these reared 
adults were checked by Dr. Alan Stone of the 
U. S. National Museum, and others were 
checked by Miss Elizabeth Marks of the 
University of Queensland. Most of the speci- 
mens have been deposited in the U. S. 
National Museum, and representative series 
of associated larvae, pupae, and adults have 
been placed in the collections of Cornell 
University and the University of Queensland, 
Brisbane, Australia. 
ANALYSIS OF DATA 
Habitats 
After all recorded data on the ecology of 
the larvae were assembled, it became obvious 
that the literature was markedly deficient in 
quantitative data. Scarcely a reference presents 
numerical data which show the actual degree 
of utilization of a habitat. Some authors use 
such terms as "sometimes,” "rarely,” "fre- 
quently,” but most merely list all habitats 
without qualification, so that it is impossible 
to distinguish the usual from the unusual. 
Undoubtedly, many authors unduly stress 
rare habitats. It seems that single collections 
from such breeding places are so repeated in 
the literature that gradually the focus is con- 
centrated on the rare to the point of neglect- 
ing the really important— the usual. 
Therefore, I believe that only a numerical 
presentation of data can furnish both a brief 
and an accurate basis for determining habitat 
potentials, and only the use of numbers will 
permit a comparison of the results of different 
studies. 
Larval habitats have been variously re- 
corded by nearly every author who has at- 
tempted to summarize an extensive number 
of collections. The majority simply give the 
range of habitats applicable to the particular 
area, listing every specific habitat, such as 
"K ration box floating in a river.” Such de- 
tails are meaningless when data are compared. 
Moreover, even though some authors at- 
tempt to classify habitats, their descriptive 
terms are unique. Thus, comparison of data 
is almost impossible. Obviously there is need 
for a classification of habitats using very 
broad terms that are universally applicable. 
Shannon (1931) proposed a very general 
classification dividing all habitats into two 
major categories: (1) depressions in the 
ground, and (2) water in containers above 
ground. Lie subdivided each of these into 
natural and artificial water. Bonne-Wepster 
and Brug (1932) found the following ade- 
quate for their investigation of Stegomyia in 
the Dutch East Indies: (1) earth-bound habi- 
tats, (2) vegetable breeding places, (3) man- 
made containers. Bates (1949) divided larval 
habitats into four major groups: (1) perma- 
nent or semipermanent standing water, (2) 
running water, (3) transient ground pools, 
(4) container habitats. Arnett (1950) sum- 
marized the habitats recognized in Panama as 
follows: (1) surface water, (2) aerial habitats, 
(3) artificial containers. 
I believe that general use of such major 
categories by all workers, supported by nu- 
merical data, would quickly give an adequate 
picture of the habitat potentials of the various 
species. Such a picture could then be utilized 
as the basis of sound control procedures. 
My organization of larval habitats cor- 
responds to that of Bonne-Wepster and 
Brug and of Shannon. They were grouped 
into three major divisions: (1) fresh surface 
water, both natural and artificial, (2) fresh 
water in containers, both natural and arti- 
ficial, (3) brackish or salt water. These were 
further divided into a rather large number of 
