Cambridge CB2 1TQ, United Kingdom
Galton studied mathematics at Trinity College, Cambridge.
Galton in his later years
Sir Francis Galton by Charles Wellington Furse, the National Portrait Gallery, London
Sir Francis Galton, 1890s
Strand, London WC2R 2LS, United Kingdom
Galton attended King’s College in 1839 - 1840.
As a child, Galton was extremely intelligent. His twelve-year-old sister, Adele suffered from a weak spine but taught him Latin and Greek from her bed anyway. She also taught him how to read advanced literature, and as he grew older, his mother would hear him reciting from Chevy Chase or Hudibras.
Both of his parents wanted him to become a medical doctor. He studied medicine at Birmingham's General Hospital and later at King's College in London. In 1840, he decided to study mathematics at Cambridge University instead of medicine.
After Galton's father passed away, he received a wealthy inheritance. He decided to abandon his studies and traveled around the British Isles and even went to Egypt and Sudan.
Even before going to Cambridge, Galton had taken an extended trip down the Danube and on to Smyrna, which had perhaps awakened the young man to the delights of foreign scenes and strange peoples. It was not until after four years of idleness in England that he set out on a trip to tropical Africa, the results of which showed that he had come to terms with life and with himself. In 1850 Galton set off for the Cape and spent two years upcountry exploring from Walvis Bay to Lake Ngami, the territory of which little was known.
He composed 15 brief laws for the Hottentot chiefs who governed the Damaras of the plain and compiled a rudimentary dictionary for the English who wished to use the local tongue. He returned to England early in 1852 and read a paper to the Royal Geographical Society.
Galton’s first publication was on exploration, and in 1855 The Art of Travel was published. In 1870, he read a paper at the British Association entitled "Barometric Predictions of Weather," in which he was fumbling toward a multiple regression, trying to predict the wind from pressure, temperature, and humidity. He failed in his objective at the time, but he posed the problem for others who were to succeed.
In assessing the intellectual influences on Galton, continuing uncertainty exists as to the extent of Quetelet’s influence. Galton’s obsession with the normal curve of error which, to a certain extent, has unduly influenced the development of statistical method, can only have stemmed from Quetelet.
The other great influence on Galton during the period in which he was establishing himself as a research worker affected the whole of the scientific world in the second half of the nineteenth century - the publication in 1859 of Charles Darwin’s The Origin of Species. The effect of this work on Galton was not immediately apparent in his writings. He began with the article “Hereditary Talent and Character” in 1865 and proceeded through Hereditary Genius (1869); English Men of Science: Their Nature and Nurture (1874); Inquiries Into Human Faculty (1883); and Natural Inheritance (1889), by which time he was 67 years of age.
In 1876, at the exhibition of scientific instruments at South Kensington, Galton exhibited his “Whistles for Determining the Upper Limits of Audible Sounds in Different Persons.” Both before and after this time he was active in proposing tests for the measurement of muscular sensitivity by weight discrimination, for the perception of differences of tint, for reaction time, for acuteness of hearing, for keenness of vision and judgment of length by the eye, and for the senses of smell and touch.
Francis Galton was associated with the American psychologist James McKeen Cattell, who on his return to the University of Pennsylvania (and later at Columbia University), began to teach statistical psychology, giving his first course in 1887.
When, however, he had determined that he had what would now be termed linearity of regression and homoscedasticity in the arrays of the table, his mathematical powers were not sufficient to enable him to form a mathematical model for his surface, and he took the problem to Hamilton Dickson, a Cambridge mathematician. Dickson’s mathematical formulation was published in an appendix to Galton’s paper “Family Likeness in Stature,” presented to the Royal Society in 1886. Galton was troubled by the fact that the slope of the regression line depends on the variability of the margins, and this concern led to his search for a unit-free measure of association.
The Bertillon system of measurement also started Galton wondering about the whole procedure of personal identification. In the paper for the Royal Institution in which he discussed bertillonage, he also drew incidental attention to fingerprints.
In his book Finger Prints (1892), he referred to the work of Jan Purkinje, Kollman, William Herschel, and Henry Faulds, who had preceded him in this study, but it is clear that at the time he wrote little was known. As a result of Galton’s book and his evidence to a committee set up by the Home Office in 1893, a fingerprint department was established, the forerunner of many such throughout the world. Galton himself, as might be expected from his previous work and interest, turned to studying the inheritance of fingerprints, a study which was carried on for many years in the laboratory that he founded and that was named after him.
The term "eugenics" was introduced by Galton in his book Inquiries Into Human Faculty (1883) and soon won general acceptance. He did more than lecture, however. In 1904 he founded a research fellowship in national eugenics at the University of London which was to develop in a few years into the Galton Laboratory of National Eugenics, with Karl Pearson as its first director. Pearson was succeeded by R. A. Fisher, and the now vast complex of statistical theory and method developed there thus owes its origin to Galton.
It was inevitable that Galton’s work should attract the interest of young men able in the mathematical and in the biological fields, and the late 1880s saw Karl Pearson, and W. F. R. Weldon - the one a professor of applied mathematics and the other a professor of zoology and both at University College, London - working in the field of “biometry,” i.e., the application of mathematics to problems of biological inheritance.
In the last decade of his life, Galton played the part of the counselor and adviser to the younger men, but he still worked away at his own problems, as his continued output of letters and papers indicates.
Galton’s first piece of fruitful research was on the weather. He started to plot wind and pressure maps and noted, from very scanty data, that centers of high pressure are associated with clockwise directions of winds around the calm center. He coined the name “anticyclone” for such systems in 1863.
Galton tried to determine a linear prediction formula for the velocity of the wind, given the pressure, temperature, and humidity. He did not succeed, possibly because of his failure to realize that the prediction formula for pressure from velocity was not the same as the prediction formula for velocity from pressure. The realization that there are two regression lines was still in the future, as was the concept of correlation.
In an attempt to describe the skewed distributions that often resulted from the application of his tests, Galton hypothesized that in some frequency distributions, such as, for example, judgment of length, the geometric mean, rather than the arithmetic mean, is the best "medium" for the distribution, and he wrote a paper on “The Geometric Mean in Vital and Social Statistics” (1879). As usual the mathematical conceptualization was beyond him, and he took the problem to Sir Donald Macalister, who derived what is now known as the log-normal distribution.
From the statistical point of view, Natural Inheritance is probably the most important of Galton’s writings. As can be seen from his earlier works, the ideas in it had been fermenting in his mind for some time, but it was their expression in Natural Inheritance that excited the interest of those whom today we might call the practitioners of applied mathematics. Again he was influenced by the fact that Quetelet was using the normal curve to describe anthropometric data and by the interest in the problems of inheritance aroused in him by The Origin of Species. He began the book with a summary of those properties of the normal curve that appealed to him.
He had previously suggested representing a frequency distribution by using grades or percentiles, and he elaborated on this suggestion here, pointing out that the normal distribution is completely determined from a knowledge of the median and one other quantile. Galton had observed that many measured characteristics can be closely described by a normal curve. He used the “quincunx,” first shown in print with the publication of his lecture "Typical Laws of Heredity," delivered at the Royal Institution (1877), to illustrate the build-up of the normal curve. In this paper, he had almost reached the concepts of both regression and correlation but must have felt the need for further thought, since it was at this time that he began to collect data bearing on inheritance in man.
Galton published nothing further on heredity for eight years. The foundation of his ideas on regression and correlation did not perhaps become clear to him until a short time before the publication of Natural Inheritance.
The regression line arose naturally out of measurements of the sizes of the seeds of mother and daughter sweet pea plants. The sizes of the seeds of daughter plants appeared to “revert” to the mean (the word “revert” was soon replaced by "regress"). This inspired him to look at a bivariate frequency table of the heights of fathers and sons, in which he found a regression to "mediocrity." The arguments he used became familiar ones with the analysis of variance put forward by R. A. Fisher some forty to fifty years later.
While studying the bivariate frequency table of heights of fathers and sons, Galton was struck by the observation that the contours of equal frequency in the table were similar and similarly situated ellipses. He also found the lines that fitted the medians of the arrays (possibly drawing them by eye) and the slopes of these lines eventually became his regression coefficients. This early work, as is inevitable with a pioneering effort, is confused and difficult to evaluate, not least because Galton himself was not explicit.
Some time earlier, in 1882, Alphonse Bertillon had put forward a scheme for classifying criminals according to 12 physical measurements that was adopted by the prefecture of police in Paris. Galton became interested in this scheme and pondered for some time over which measurements would be the most descriptive - that is, which would discriminate one man most effectively from his fellows. It was from these considerations that he was led to the realization that some measurements might be so highly correlated with other measurements as to be useless for the prescribed purposes and finally to the necessity for describing how any two measurements are related.
The slope of the regression line is not adequate for this since it depends on both the scales of measurement and the choice of dependent variables. However, the regression line fitted between the variables that Galton used (1888) after dividing the heights (reduced by their median) by a measure of their variability (their semi-interquartile range) and similarly dealing with forearm length provides a unit-free measure of association. Given the problem and 65 years of subsequent statistical development, the correlation coefficient may now appear to have been inevitable.
"What nature does blindly, slowly and ruthlessly, man may do providently, quickly, and kindly. As it lies within his power, so it becomes his duty to work in that direction."
"The phrase 'nature and nurture' is a convenient jingle of words, for it separates under two distinct heads the innumerable elements of which personality is composed. Nature is all that a man brings with himself into the world; nurture is every influence without that affects him after his birth."
"Eugenics is the study of the agencies under social control that may improve or impair the racial qualities of future generations either physically or mentally."
"I HAVE no patience with the hypothesis occasionally expressed, and often implied, especially in tales written to teach children to be good, that babies are born pretty much alike, and that the sole agencies in creating differences between boy and boy, and man and man, are steady application and moral effort. It is in the most unqualified manner that I object to pretensions of natural equality. The experiences of the nursery, the school, the University, and of professional careers, are a chain of proofs to the contrary."
"One of the effects of civilization is to diminish the rigour of the application of the law of natural selection. It preserves weakly lives that would have perished in barbarous lands."
"The aim of eugenics is to represent each class or sect by its best specimens; that done, to leave them to work out their common civilization in their own way."
Francis Galton was elected a member of the Athenaeum Club in 1855 and made a Fellow of the Royal Society in 1860. He was also one of the founders of the Lunar Society of Birmingham.
The range of questions Sir Galton devoted his time to was extremely wide. He was a very intelligent person, which allowed him to make a serious contribution to many fields of science.
Galton married Louisa Jane Butler, on August 1, 1853. The union of 43 years proved childless.