1968] 
Coyle- — Spider Genus Atypoides 
159 
characters are given in the appendix at the end of this paper.) 
These data were analyzed with the aid of a 7094 IBM computer at 
the Harvard Computer Center. The computer program, written in 
Fortran II, calculated the mean and standard deviation of each 
measurement, meristic character, and each of 44 different ratios 
formed from these for each population sample of each sex and for 
certain groupings of samples plus individual specimens into infra- 
specific or species units. The program then compared these samples 
and groupings pairwise in all desired combinations giving for each 
character for each comparison a value of the distinctness of the two 
samples. This value, called the “distance”, is equal to the difference 
between the mean of each of the two samples divided by the sum 
of their standard deviations. This enabled me to quickly select 
those characters of greatest diagnostic value, those characters which 
show the most marked geographic variation, and those infraspecific 
samples that were most divergent. 
As will be evident in the species diagnoses and Tables I and II, 
many of the measurements, counts and ratios are diagnostically use- 
ful in this genus. Several measurements and counts which are not 
diagnostically useful are included in the tables for their descriptive 
value or because they are diagnostically important in Antrodiaetus , 
a genus in which the species are often morphologically much more 
similar to one another than is the case in Atypoides. The analysis 
of geographic variation is considered preliminary because of the 
small sample sizes and the lack of more samples from important 
parts of each species’ distributional range. 
Each of the localities from which a population sample of A. 
riversi or A. gertschi was obtained is labeled on Map I and identi- 
fied in the locality records by a capital letter. These letters will be 
used throughout the text when referring to a particular sample or 
locality. The population sample sizes are indicated in the Dice-Leraas 
diagrams and the species sample sizes are given in Tables I and II. 
Measurements. All measurements were performed by myself 
with the same binocular stereomicroscope and eyepiece micrometer 
scale. A series of three specimens was remeasured for each char- 
acter five times during the course of the study and indicated that the 
measurements were accurate to one micrometer unit for each of the 
four different powers of magnification used. One micrometer unit 
had the following value for the following characters: 0.0753 mm 
for CL; 0.0377 mm for CW, SL, SW } and all leg and pedipalp 
segment lengths; 0.0182 mm for PTT; and 0.0092 mm for CAT 
and all eye measurements. 
