SummaryPart 1Frequency distributions for the finger ridge‐count on each of the ten digits in a population sample of 825 males and 825 females are given, together with their respective means and standard deviations.The highest means are those for digit I, followed in order by those for digits IV, V, III and II.The mean ridge‐count is significantly greater on each digit in males, the difference being greatest on digit I of both hands.The mean ridge‐count on the right hand is significantly greater than that on the left on digit I in both sexes and on digit II in females.Digit II shows the greatest variability in ridge‐count and digit V the least. Part 2A value S2, the sum of squares of deviations of the ten separate finger ridge‐counts from their mean, has been used as a measure of the variability between the digits of an individual.The joint distribution of this variable with the total finger ridge‐count, Q, is given for 825 males and 825 females, respectively.In both sexes the distribution is arch‐shaped, with entries decreasing from the base to the apex, while the regression of S2 on Q is a curved line. In each case there are signs of trimodality.The joint distribution of (S/√ 10) and Q is also given for each sex.The correlation between S2 and Q is negative in both sexes: r=–0–195 ± 0–03 for males and r=—0–12 + 0–03 for females. These values are significant.The characteristic features of the joint distribution of S2 and Q are discussed with a view to their possible genetic significance.The data used for this paper are filed in the archives at the Galton Laboratory.