francis galton count and measure

Upload: anca-popescu

Post on 06-Apr-2018

240 views

Category:

Documents


1 download

TRANSCRIPT

  • 8/3/2019 Francis Galton Count and Measure

    1/4

    Copyright0 1993 bv the Gene tics Society of America

    PerspectivesAnecdotal, Historical and Critical Commentaries on Genetics

    Edited by James F. Crow and Wil l iam F. Dove

    FRANCIS ALTON:ount and Measure, Measure and CountJames F. Crow

    Genetics Department, University of Wisconsin, Madison, Wisconsin 53706

    WOULD DARWINave understood and appreci-ated MENDELSwork had he known of it ? Thequestion is often asked. B u t DARWINas so convincedof the continuous nature of inheritance that hemightnot have been receptive. Ther e is a better candidate,FRANCIS ALTON.n 1965 A. H. STURTEVANTrote:The question has often been raised: Would any biologisthave appreciated Mendels work f he had seen the paperbefore 1900! My own can didate for the most likely personto have understood it is Galton, because of his interest indiscontinuous variation, his mathematical turn o f m ind, andhis acceptance of Weismanns view that the hereditary po-tentialities of an individual must be halved in each germcell.GALTONived long enough to learn about the redis-covery of MENDELS aws, but by this time his geneticwork was largely finished.

    GALTONwas the founder of biometry. Regressionand correlation were his ideas and he used them tostudy inheritance and the similarity of relatives. Hepioneered in the use of the Gaussian distribution, andrealized that the methods developed to study meas-urement errors were just what was needed to studybiological variability. He showed that one measuredtrait after ano ther was normally distributed, or closeto it. He did monumental studies f stature, eye color,and hereditary disease in humans. From the extensiverecords of coat color in basset hounds he formulatedhis law of ancestral heredity, nice idea butsuperseded by Mendelism. He was the first to usetwins to separate genetic from environmental ffects.Along the way, he developed techniques forlassifyingfingerprints, nteresting to him asaneutral rait.GALTONSigh place in the history of genetics is shownby a biographical memoir and portrait (Figure 1 ) inthe second volume of GENETICS,ollowing MENDELwho was honored in the first.

    GALTONSreat work onheredity and biometrycame late in his ife. He published his Hereditary Geniuswhen he was nearly 50 and Natural Inheritance whenGenetics 135: 1-4 (September. 1993)

    FIGURE --FRANCIS GALTON, from a painting by C. Mr. URSE.Reproduced from the frontispiece of GE NE PI CS, Volume 2, I9 17 .he was well into his 60s. Th e last third of his long lifewas spent almost entirely on heredity, biometry, andeugenics, and it is these for which he is known. Yethis earlier life and accomplishments are also remark-able, although lesswell known to geneticists, and Iwill mention some of them. Much of this comes fromhis fascinating autobiography, Memories of My Life,published in his 88th year.

    I have also been dipping nto a four-volume, 15-pound biography comprising 1345 pages, plus fold-outs and page after page of photographs. I t was writ-ten by KARL PEARSON (191 4, 1924, 1930a,b).n thepreface PEARSONays, T he indolent reader will findmuch in this work which he does not want and whichis of little interest to him . . . this work is not writtento gain a public. And in Volume 111 he says further,

  • 8/3/2019 Francis Galton Count and Measure

    2/4

    2 J. F. Crow I will paint my portrait of a size and colouring toplease myself, and disregard a t each stage circulation,sale or profit. As the publishers later rued, Thisquaint notion of the authors, which some would callintegrity, resulted in a rather small sale. Th e over-stocked vender finally offered the ntire set for$17.50. A set now sells for several hundred dollars.

    GALTON1822-191 1) was born the same year asMENDEL (1 822-1 884). DARWIN 1809-1882) andGALTONwerehalf-cousins, grandsons of ERASMUSDARWIN.GALTON as the youngest of nine childrenand was carefully attended by a sister. She had a spinaldeformity and was confined to her bed, from whichshe directed his early studies. He was a precociouschild. On his fifth birthday, uninhibited by false mod-esty, he wrote, I am four years old (sic) and I canread any English book. can say all the Latin Substan-tives and Adjectives and active verbs besides 52 linesof Latin poetry. I can cast up any sum in addition andcan multiply by 2, 3, 4, 5,6, 7 , 8, [9], 10, [111. I canalso say the pence table. I read French a little and 1know the Clock. The numbers in brackets wereerased from the letter; inhis exuberance he hadclaimed too much. In school he led his class, althoughthere were comments, then and later, that he hadneglected his classics in favor of natural history andscience. His intense curiosity about everything and hisdetermination to experiment were apparent atanearly age and continued throughout his life. His con-viction that counting and measurement were the se-cret of success grew with the years.

    At his parents urging he began medical studies atage 16. He wrote feelingly about the brutalities ofsurgery-this was before anesthetics-and wonderedwhy more patients were not made dead drunk beforesurgery, which n one instance seemed to providesuccor. Always the experimentalist, he decided to findout for himself the effects of all the medicines bytaking very small doses. Going through the list alpha-betically, he reached the letter C; but even a minutedose of croton oilwas too much for him. He alsoflirted with hypnotizing people and discovered, con-trary to the conventional wisdom of the time, thatintense concentration on his part was unnecessary; hewas just as effective when he deliberately let his mindwander.

    GALTONwas admitted toTrinity College,Cam-bridge, and started to concentrate on mathematics. Ahigh point was a study-vacation tour with mathematicstaught by ARTHURCAYLEY,nventor of matrix alge-bra. Unfortunately, GALTONSlans for further studywere halted by illness. Later when he returned to hisstudies, he gave up his mathematical ambitions andsettled for a medical degree. At about this time hisfather died, leaving him with a fortune sufficient to

    make medical practice unnecessary, and immediatelyhe abandoned it .The next few years were spent traveling, mostly in

    Africa. He rode on horses, oxen and camels, and ona barge up the Nile.His trips included areas notpreviously traveled by Europeans. Everywherehewent he made notes and measurements. Latitude andlongitude were determined by astronomy and altitudeby the temperatureof boiling water. He wrote severalbooks about traveling, and for many years was activein the Royal Geographical Society.Always,even on trips, he read widely. He wrotefrequent articles, often for Nature, on all manner ofsubjects. He also was an inventor. He invented a newtype of heliostat, for sending signals by sun reflection.He made a trip to Spain to study a total eclipse of thesun with a newly invented instrument for measuringtemperature changes; it failed, but this, he said, gavehim time to revel in the beauty of the eclipse. Hedeveloped stereoscopic maps, better to show relief.He experimented with trying to make breathing en-tirely voluntary, and almost succeeded, to his terror.

    He became interested in the ability of animals tohear frequencies too high for human ears and de-signed supersonic whistles to measure this. In orderto measure the hearing of zoo animals without beingnoticed by human visitors, he contrived a whistle inhis cane that he could unobtrusively activaterom thehandle. T he results, however, were disappointing.He wondered why, with all the selection in dogs forspecial traits and bizarre shapes, no one had everselected for intelligence; but he did not have theresources for a study and could not elicit interest fromanyone else. He studied a calculatingprodigy andlearned about the geometrical pattern in which thisperson arrayed numbers in his mind.

    GALTONSuriosity and efforts to satisfy itexceededall bounds. He once decided to study the efficacy ofprayer. To do this he studied mortality rates of roy-alty, whose subjects prayed for their health, and com-pared insurance charges for vessels with and without(presumably praying) missionaries.

    Another of GALTONSnventions was to combinephotographs of different individuals into a singleprint. In this way he could wash out he noiseofspecial features and reinforce the general ones. Thisenabled him more readily to perceive amilial andethnic resemblances. As usual, he invented a machineto make the composites. This foreshadowed moderncomputer programs for removing noise from images.

    This superficial summary may give the impressionthat these various experiments were done casually,which is not true. Almost always hey involvedcareful,repeated measurements, and were usually published.GALTONSard-working habits paid off ina list of 183articles and memoirs, including 12 books.

  • 8/3/2019 Francis Galton Count and Measure

    3/4

    Perspectives 3GALTONSmost amous experiment is his est ofDARWINSangenesis hypothesis. In the later ditionsofhisbook DARWIN ostulated that minute gem-

    mules pass from body cells to the germ cells, thusproviding a mechanism for the Lamarckian inherit-ance that he was increasingly drawn to. GALTONestedthis bymassive transfusions in rabbits. The experi-ments involved cross onnections of the blood vessels,leading to a thorough mixing. He praised the skill ofhis friends who did the tricky surgery; his own partwas confined to inserting cannulae and he like.(GALTONlways seemed to have friends willing to helphim.) The results were unequivocal: there were noinherited effects of the transfusions. DARWIN as notconvinced. He said that gemmules need not travel bywayof the blood stream and hat his theory wassupposed to apply as well o organisms without blood.GALTON as himself argely convinced of the non-existence,or at least the non-importance, of Lamarck-ian inheritance, although, perhaps out of respect forhis cousin, he kept an open mind. He did, however,retain part of the pangenesis theory, but confined theparticles to the germ line. If only he had understoodMendelian segregation and recombination!Another of GALTONSasting contributions was thedevelopment of fingerprints. At this time BERTILLONwas using body measurements as a way of identifyingpeople. GALTON ointed out a fundamental er ror,that the various measurements were not independentand the simple multiplication of probabilities couldgive an erroneous impression of accuracy of identifi-cation, the same issue that sometimes arises today inDNA forensics. Looking for something better, GAL-TON had the idea of trying fingerprints, which hadbeen usedby, among others, the ancient Chinese.GALTONSreat contribution was developing a system-atic way of cataloging fingerprints to permit readycomparison. His methods and derivatives therefromare now standard. He was fascinated by this traitbecause of its apparent selective neutrality. How, hewondered, could such seemingly useless patterns bes o precisely determined, as evidenced by the regen-eration of the same pattern when the finger tip hadbeen mutilated? GALTON adoriginally thought offingerprints as a way to aid racial classification. Theyturned out o be useless for this purpose, but far moreimportant as a means of individual dentification.Although GALTONid all of hese things, and muchmore, his deep interest in the last part of his life washeredity. Increasingly he emphasized counting andmeasurement. His greatest accomplishmentswere thediscovery of regression and correlation, the founda-tions of biometry. Filial regression is now so familiarthat it is hard to imagine that its significance was soslow to be appreciated. GALTONoted that the prog-eny or other relativesof ndividualswho deviated

    from the population mean n either direction alsodeviated in the same direction but to a lesser extent. Ifthe parental measurement deviated by an amount x,say, the progeny deviated by an amount Kx (0 < k

  • 8/3/2019 Francis Galton Count and Measure

    4/4

    4 J. F.Crow

    FIGURE .-A quincunx. In this example a triangular block hasbeen introduced to produce a skewed distribution. (Reprinted bypermission of the Quantum Company.)England. From the large number of outstanding peo-ple (e.g., Plato, Socrates, Sophocles, Arisophanes) n asmall population he calculated that the average apti-tude in ancient Greece was about 2 units (1.35 stand-ard deviations) above that of contemporary Britain.Viewed by 1993 standards this seems looseand naive.But it illustrates beautifully GALTON'Sbsession withmeasuring everything and usingwhatever rick hecould think of.

    One of GALTON'S ost appealing inventions was adevice for dropping pellets through a hole, wherethey bounced off strategically located pinsFigure 2).A t the bottom they accumulated in slots and gener-ated a binomial distribution, which became Gaussianas the number of pelletsand slots increased. He calledthis a quincunx after the geometrical pattern of thepins. By suitable modifications he could demonstrateselection, regression, and various statistical principles.The most ingenious was his use of a two-stage quin-cunx to demonstrate that the sum of a number ofnormal distributions is itself normal.

    GALTON'Selightful account ofhow he came toinvent the concept of correlation was published in1890 and has been reprinted with an introduction bySTICLER1989). For a discerning discussion of GAL-TON'S central role in the history of statistics see STIG-LER (1986).

    GALTONelieved firmly in the importance of hered-ity, and reacted strongly against the prevailing viewthat mental traits were not inherited. "I have nopatience with the hypothesis occasionally expressed,and often implied, especially in tales written to teachchildren to be good, that babies are born pretty muchalike, and that the sole agencies in creating differ-ences . . .are steady application and moral effort. Itis n the mostunqualified manner that I object topretensions of natural equality." His books are full ofstatements that are today regarded as excessivelyher-editarian or racist. There is a cottage industry of booksand articles critical of this. But I say it is unfair tojudge a Victorian Britisher, a creature of his time, bypolitical and social standards of the 1990s. I prefer toadmire him for his achievements. he subject of biom-etry is a lasting monument, unaffected by the chang-ing winds of political orrectness.

    My greatest debt is to STEVE TICLER ho has made a detailedstudy Of GALTON.also thank CARTER ENNISTONnd BILLENGELSfor helpful comments on he manuscript and BRIAN OINER orhelping me find a quincunx.

    L I T E R A T U R E C I T E DFISHER,R. A. , 1918 Th e correlation between relatives on the

    supposition of Mendelian inheritance. Trans. R. Soc. Edinb.GALTON, ., 1869 Hereditary Genius, Ed. 2, 1892. Mamillan, Lon-GALTON, ., 1889 Natural Inheritance. Macmillan, London.GALTON, ., 1909 Memosics of M y Lqe. Menthuen, London. (This

    includes a nearly complete list of GALTON'S ublications, buteven GALTON imself could not keep track of all that hewrote.)PEARSON,., 1914, 1924, 1930a. 1930b Lqe, Letters, and Laboursof Francis Galton. Th e University Press, Cambridge.PROVINE, . B., 1971 TheOrigins of Theoretical Population Ge-netics. University of Chicago Press, Chicago.STICLER, . M., 1986 The Hisfory of Statistics. Harvard UniversityPress, Cambridge.STIGLER,. M., 1989 Francis Galton's account of the invention ofcorrelation. Statist. Sci.4: 73-86.STURTEVANT,. H., 1965 A Histoy of Genetics. Harper 8c Row,New York.YULE,G. U., 1902 Mendel's laws and their probable relations tointra-racial heredity. N ew Phytol. 1: 193-207.222-238.

    52: 399-433.don.