FEEDING TIMES 
177 
3. Numbers of Euschongastia (Walchiella) oudemansi recovered day by day from a single 
wild-caught rat. 
The chiggers clearly detach about two or three modal times. These probably 
represent two or three separate mass infestations. 
It seems fair to assume that any rat so qualifying had not yet lost the modal group of its last 
infestation. The maxima therefore should range from the value by which the maximum 
exceeds the mode to the true maximum, and the mean value of all maxima should approximate 
to the true maximum less half the mode. 
2. Angular transformation. — If all apparent feeding times for any one species of mite are 
summed a sigmoidal curve is obtained, such as that shown in fig. 4 in which the total numbers 
of Trombicula deliensis , from twenty-four wild-caught rats, which had been recovered by any 
one day are shown as a percentage of the total number recovered, on a logarithmic scale of time. 
If the number of mites attacking a population of rats each day is approximately constant, then 
the population of mites on the rats will clearly settle down to some constant number, N say, 
from which a constant number, n say, leaves daily, and to which the same number, n, is added 
daily. The mean feeding time is clearly ^ which is the inverse of^the proportion leaving daily. 
If the whole population of rats is now trapped the number of mites leaving daily should be n, 
but since the mite population is not replenished this number gets less as time goes on. The 
problem of estimating the mean feeding time can therefore be resolved into one of estimating 
the proportion of mites on a population of rats which come off the rats during the first day 
after trapping. The inverse of this is the mean feeding time. 
From data of the sort shown in fig. 4 this number could be estimated by fitting a curve to 
the points and seeing where it cuts the vertical equivalent to day one. The curve however 
MALAYA , No. 26, 1953 
