4 
SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 1 52 
SAMPLE PAIRS 
Seven pairs of samples were taken in April 1965. Each pair was 
taken from a single grab sample 0.1 sq m in area. To examine homo- 
geneity of species proportions at a station (not between stations), the 
data were arranged in contingency tables. The “goodness of fit” 
between the observed and expected numbers was then analyzed by 
means of chi-square. Species for which the expected frequency is less 
than two were deleted from the calculation of chi-square. For some 
pairs (sta. 262, 265, 266) the expected frequency for all species is 
less than two and no analysis was attempted. These contingency 
tables are shown in table 2. 
Of the pairs which could be tested, none of two for the living popu- 
lation and two of four for the total population have species propor- 
tions which agree with expectation. Although the pair 262-262' could 
not be tested, it is clearly heterogeneous. The pair 266-266' and the 
living population of the pair 270-270' may also be heterogeneous. In 
the Complex as a whole we may be confident that species proportions 
are homogeneous in sample pairs at about one third of the stations. 
In this respect the Hadley Harbor Complex is similar to the nearshore 
area of Long Island Sound. Buzas (1965) found two of five sample 
pairs homogeneous for the living population and one of six for the 
total population in the nearshore (<20 m) areas of Long Island 
Sound. The offshore areas of Long Island Sound are, however, more 
homogeneous than the nearshore areas. In the Hadley Harbor Com- 
plex, sampling is not adequate to permit drawing any conclusion 
between subareas. 
The reason why sample pairs often fail a test for homogeneity is 
because individual foraminifers are not all randomly distributed. 
Buzas (1968) has shown that species with relatively low densities 
have randomly distributed individuals, but as the density increases 
they become progressively more aggregated. This is an unfortunate 
situation because those species which are most abundant and conse- 
quently most likely to be included in any test of homogeneity are the 
ones which are most aggregated in their distribution. 
FAUNAL COMPOSITION 
Sixteen species were recorded from the samples taken in July 1964 
(see table 1). The most abundant species was Elphidium clavatum. 
E. incertum , Buccella frigida , and Ammonia beccarii were also rela- 
tively abundant. A relatively large number of individuals belonging 
to Quinqueloculina seminula were found at stations 96 and 107. 
