DECLINE IN FREQUENCY OF TUNNEL SEGMENTS OF EVER INCREASING 
LENGTH ( 3-POINT MOVING AVERAGE O) 
O 150 350 550 750 950 1150 1350 
LENGTH OF TUNNEL SEGMENT IN 
MILLIMETERS (d ) 
Figure 45. — Lengths of tunnel segments — 3-point moving 
average. Mote that the frequency of tunnel segments 
declines more rapidly than does an inverse proportionality 
to distance. 
tunnel segment, the proportionately greater will 
be the work involved. In summary: 
Type of work 
1st term: C= decision 
2d term: CX— digging from end of tunnel 
CX 2 
3d term: — — transportation 
y fZ 
Later terms, to C— y will presumably involve 
other behaviors such as the likelihood of the dirt 
carried in the mouth crumbling, stopping to defe- 
cate, et cetera. The farther the tunnel is extended 
from the point of decision, the more behaviors 
come into play. The relative influence of each of 
these behaviors (i.e., sequential terms in the 
equation) may be seen diagrammatically in figure 
47. 
At any d, the effort equals the sum of the values 
X n 
of the curves for the factors C to C— r • 
n\ 
That is to say, at the shorter distances the effort 
is mainly determined by the decision (C) and 
digging (CX) terms. For the longer length 
tunnels the later behaviors ^terms, C^-^ . . . C ^ ^ 
make a relatively greater contribution to the total 
work involved. I am indebted to Mr. James U. 
Casby (formerly of the Army Medical Service 
Graduate School) for the concepts presented in 
this section on the “Pattern of work as reflected in 
burrow construction.” 
J. The Location of Chambers. At the end of some 
tunnels rats make enlargements which may be 
used for either nests or storing food. Distances 
from the center of chambers to the outside were 
taken for 178 chambers from the 44 burrows 
(fig. 48). The half of the records involved in the 
shorter distances appear to be approaching an 
asymptote. Some factor existing outside the 
burrow must be sufficiently ameliorated by the 
time a distance of 550 mm. is attained that the 
internal environment of the burrow is at an opti- 
mum. Beyont this point the character of the 
distribution abruptly changes to one of an exponen- 
tial equation governing the lengths of individual 
tunnel segments. This section of the curve is 
probably a direct reflection of the fact that where 
one tunnel segment branches off from another the 
laws governing the frequency of the sum of the 
lengths of two or more tunnel segments follows 
closely those governing the frequency of single 
tunnel segments. The reason for believing this 
is that slightly more than 50 percent of the chambers 
occur at the termination of two or more tunnel 
segments from the outside. An analysis of four 
of the larger burrows gave the following: 
Number of tunnel segments between 
chamber and nearest exit to the outside . . 1 2 3 4 
Number of chambers 28 30 6 4 
K. The Balance Between Types of Tunnel Segments , 
and its Relationship to Social Organization. Three 
types of tunnel segments may be discerned: (a) 
Blind. These include all those whose terminal 
end is without bifurcation or enlargement into 
chambers, (b) Exit. These include all those 
which connect on one end with an opening to the 
surface, (c) Internuncial. These include all those 
which do not connect with an opening to the 
surface but which have either a chamber or another 
tunnel segment at either end. 
The proportion of the total formed by each of 
these three types of tunnels changes as the size of 
the burrow increases (see fig. 49). While the 
burrow is in its initial stage of expansion, up to 10 
tunnel segments, there is an increase in the blind 
tunnels until a stable level of approximately 15 
percent is reached. The exit tunnels gradually 
46 
