In the fall of 1909 Morgan began to experiment on the variable trident-shaped pattern on the thorax of Drosophila, selecting and inbreeding stocks in which the trident marking was least and most pronounced. This kind of "pure-line" selection was the commonest variety of experimental evolution at the time. Neo-Darwinians like Castle used it to model in the lab what they believed to occur as species evolved in nature. Anti-Darwinians like Wilhelmjohannsen used the same experiments to show that selection of minute variations could not result in permanent change. So what did Morgan hope to show with his selection of the trident marking? Certainly he was not emulating Castle; nor, I think, was he recapitulating Johannsen's classic work.

The timing of the trident experiments, coming not long after Lutz's discovery of wing-vein and dwarf mutants, suggests that Morgan was trying to reproduce Lutz's work and to explore the possibility, which Lutz had not, that the experimental process of selection had itself produced mutations. [1] This is supposition, but it is consistent with Morgan's customary working habits, and we do know that he and Lutz were exchanging Drosophila stocks and lore at the time. In the summer of 1908, just after his wing mutants turned up, Lutz went to Woods Hole especially to see Morgan about them: "I am sort of run out of inspiration," he confided to Zeleny, "and I know of no better place to get some more." [2]

But exactly what kind of mutational event did Morgan expect to see? It has usually been assumed that he was looking for what he actually found: a visible, striking change like the white-eyed mutant. But Morgan could not have been expecting to produce de Vriesian macro-mutations, because he did not believe that such large mutations gave rise to new species in nature. They were too rare and too extreme. The raw materials of evolution, in Morgan's view, were definite mutations within the range of normal variation of a species or variety. If one of these small but definite mutations could induce another to occur in the same direction, the process could, he thought, lead to the rapid evolution of new species. [3] These mutations would appear not as visible characters but as abrupt steps in the selection process. Apparently stable "pure" lines of high and low trident would suddenly be capable of further and more extreme modification by selection. That, I think, is what Morgan expected a de Vriesian mutating period to look like if it could indeed be induced in the laboratory by intense artificial selection.

The selection experiment with trident was different from Morgan's earlier experiments with chemically and physically induced mutation, in which he may well have been hoping to produce large, visible mutants. Morgan appears to have designed the trident experiment to simulate more closely the actual process of evolutionary change in the wild. He was hoping to invent a new mode of experimental evolution that was different from those of de Vries, Castle, and Cuenot precisely in dealing with the kind of variations on which, he suspected, selection acted in nature. Morgan disparaged neo-Darwinian experiments with fluctuating characters as mere "meddling" with an organism, upsetting their natural physiological balance in a way that could never occur in nature. [4] To him, selection experiments like Castle's were laboratory exercises that had nothing to do with evolution in the wild. Nor did the experiments of Neo-Mendelians like Cuenot, in his view; they might illuminate the process of hereditary transmission but were irrelevant to evolution because they dealt only with characters artificially altered by domestication. For Morgan, a true experimental evolution dealt with the kind of variations from which new species were actually created in nature: that is, mutations that were discontinuous but within the normal range of variation. [5] In other words, Morgan was using the selection method of the neo-Darwinians to detect mutations too small to show up in de Vries's approach.

Experience did not immediately conform to expectations. By inbreeding lines with light and dark trident patterns, Morgan quickly produced strains that selection could not further change. But then for months nothing happened. Morgan grew discouraged. Ross Harrison recalled visiting Morgan's lab in the first days of January 1910: "'There's two years' work wasted,' he [Morgan] exclaimed, waving his hand at rows of bottles on shelves. 'I've been breeding these flies for all that time and have got nothing out of it.'" But just a few days later, Morgan observed a few flies with a darker pattern than any he had seen before. Inbreeding quickly produced a mutant strain, with, which had a distribution of pigmentation that was distinctly higher than the wild type, and which further selection did not alter. It was just what Morgan expected the process of speciation by mutation would look like! Visiting his wife and newborn daughter in the hospital in the second week of January, Morgan could talk of nothing but his new mutant. No wonder: Morgan thought he was witnessing the beginning of a mutating period - evolution induced by strenuous selection in the laboratory, where all could witness it! [6] A second mutant, superwith, appeared in November 1910, thirty generations later.

Morgan's various efforts to induce mutation by altering the physical environment and by selection were in my view preparations for a major project in experimental evolution, parallel to regeneration, sex determination, and Mendelizing. The pattern of his behavior is unmistakable. He wrote his 1903 book and subsequent articles in 1905 and 1909 to assimilate current work on adaptation and evolution and to stake out a personal position on key issues. He did experiments on Drosophila and other insects to get hands-on experience of the ways that experimental evidence about variation was produced. This was just how he had broken into the neo-Mendelian game, criticizing Cuenot and getting a feel for mouse breeding. By 1909 he had a distinctive view of variation that served as a practical guide for breeding experiments, and he thought that Drosophila might be the ideal organism for experiments on evolution, just as mice were for Mendelizing and Phylloxera was for sex determination.

Morgan chose Drosophila for experimental evolution, rather than some domesticated plant or animal, precisely because it was not domesticated. He had long believed that experiments on evolution would have to be done with wild creatures. "Until we know more about the results when wild varieties are crossed with their wild species or with other varieties," he wrote in 1909, "we can not safely apply Mendel's laws to the process of evolution." [7] It is also clear why Morgan did not use Lutz's domesticated stocks of Drosophila but instead collected fresh material from nature. If Morgan believed that Lutz's inbred stocks had already entered into a mutating period, he would have wanted to avoid such stocks in experiments designed to see if artificial selection could induce such an event. (Morgan also insisted that Payne use a wild stock for his selection experiments, not one from Lutz, doubtless for the same reason.) [8] Drosophila's natural wildness and lack of Mendelizing characters, which so limited its use for genetic experiments, made the organism of choice for Morgan's new program in experimental evolution. Morgan meant Drosophila not to replace mice but to establish a new niche and a new line of experiment in the ecology of his lab.

The appearance of with in January 1910 was just the harbinger, as turned out, of a gathering flood of mutants: olive body color and speck wing axil in March, beaded wing and another olive in May, along with the famous white eye-color mutant. Rudimentary wing appeared in June, pink eye in July, miniature and truncate wings in August, and at least six more by the end of the year. It appeared to Morgan that his selection experiment had indeed inaugurated a de Vriesian mutating period in his flies. The character of these mutants, however; was not quite what the theory predicted. Only the olive mutants were like with, definite but within the range of normal variation. The others were more extreme, more like de Vries's Oenothera mutations of Mendelizing sports. Although they did not display precise Mendelian ratios, these mutants did segregate. And white eye, to Morgan's astonishment, was sex-linked. That is, it appeared only in males, which had therefore to have a single sex chromosome. This freshet of extreme mutations was not what Morgan had expected of a mutating period, and it entirely upset any plans for a project in experimental evolution using Drosophila.

Indeed, the freshet of mutants in Drosophila completely upset the hierarchy of experimental lines and organisms in Morgan's lab. With the appearance of white and other Mendelizing mutants, Drosophila broke out of its restricted domain of experimental evolution and began to take over Morgan's other experimental lines. The sex-linked character of white made it an ideal organism for work on sex determination, better even than Phylloxera, which was soon set aside for studies of sex chromosomes and sex determination in Drosophila. Similarly, the fact that white and other mutants displayed more or less Mendelian inheritance made Drosophila the organism of choice for experimental heredity, better than mice or rats with their limited number of characters. Mice were soon displaced from their favored niche in the Columbia lab. In short, the proliferation of mutations in Drosophila altered the domestic ecology of experimental organisms and disciplines. The fly began to take over Morgan's entire operation, displacing the standard organisms in his other lines of work. [9] Morgan's mouse colony had been little used for some years before it was finally destroyed in a fire in 1914. Drosophila had taken over its role in experimental heredity as effectively as it had displaced Phylloxera from experiments on sex determination. No creature could compete with the little fly.

Experimental evolution, ironically, was the one line that did not survive Drosophila's takeover of Morgan's lab. As it gradually became clear that the flood of mutants resulted simply from scaled-up production and not from a de Vriesian mutating period, selection experiments ceased to have any relevance to the mechanism of evolution. Besides, Morgan was not one to miss an experimental windfall, and the Mendelian analysis of mutants like white was the kind of opportunity that comes seldom in a lifetime. Morgan did continue selection experiments on with and similar mutants for a year or so, but his gradual abandonment of this line of work for neo-Mendelian experimental heredity marks the transition from experimental evolution to modern genetics.

Morgan's last brief fling with experimental evolution began in November 1912 with the appearance of streak, a slight and highly variable thorax-pattern mutant. Bridges and Sturtevant would have discarded it as useless for mapping, but since it had turned up in a stock that had never been subjected to selection, Morgan saw a chance to prove what everyone now believed, that selection did not induce mutations of the with type. He worked on streak for six months, "not very vigorously" and to little effect. The next such mutant that appeared, trefoil (November 1913) was discarded forthwith, and in June 1914 all the with stocks were thrown out and soon forgotten. [10] It is striking how completely this crucial stage in the invention of genetics was wiped from the drosophilists' collective memory. When Lillian Morgan recalled her husband's visit to the hospital in January 1910, she remembered him talking about the white mutant. It must have been with, however; since white only turned up in May!

The invasion of Drosophila mutants thus turned Morgan's orderly hierarchy of experimental organisms and disciplines topsy-turvy. Drosophila took over its new ecological niche, displacing established organisms and imposing a new order on the organization of experimental work. It would change forever the natural history of experimental biological laboratories.

New HMS Beagle Cutting Edge Dialogue on Model Systems - scientists, historians, and philosophers discuss the strengths and limitations of using Drosophila and other model organisms for developmental, genetic, and evolutionary studies.

Thomas Hunt Morgan - an essay at the Nobel Foundation Web site on the life and work of the researcher who won the 1933 Nobel Prize in physiology or medicine.

The Drosophila Virtual Library - a directory with numerous links to Drosophila melanogaster resources. Includes an introduction for newcomers, plus links to the Drosophila Genome Project, the Encylopedia of Drosophila, the European Drosophila Genome Project, Drosophila Labs on the Web, Virtual FlyLab, JFly (Japanese Drosophila resources, protocols, and much more.

Flybrain - an online atlas and database of the Drosophila nervous system. Sections include What's New in Flybrain?, Basic Atlas of the Drosophila Brain, Gene Expression, Immunocytochemistry, Terms and Abbreviations, Poster Session, Development, and Links to Allied Data.

Carthew Lab Manual - A manual of Drosophila protocols. Sections include Fly Food Recipe, Quarantine of New Stocks, Generating Somatic Clones by FRT Recombination, Microscopic Analysis of the Adult Retina, Adult Head Frozen Tissue Sections, and Immunostaining Drosophila Embryos. The site also includes plasmid maps.

Berkeley Drosophila Genome Project - consortium determining the complete DNA sequence of the fly's euchromatic genome, and developing experimental and computational tools to probe its biological significance. Links include the Encyclopaedia of Drosophila; Genomic DNA Sequences; EST Project/cDNA Sequences; CK EST Project; P Element Mutations. The BDGP is funded by and works in collaboration with the National Human Genome Research Institute.

GIFTS (Gene Interactions in the Fly Trans-World Server) - a database server for gene interactions. Sections include an introduction (in English and French), GIFTS Databases, Services, Documents, Research Information, and Postdoc Positions. Presented by the Bernard Jacq group at the Laboratoire de Genetique et Physiologie du Developpement, Marseille.

FlyBase - a database of the Drosophila genome. A comprehensive database for information on the genetics and molecular biology of Drosophila. Sections include Primary Data (Genes, Aberrations, Clones, Maps, People, Stocks, References, Images, Transposons and Vectors, and Searches), Documents, Nomenclatures and News; Drosophila Genome Projects, and Allied and Other Databases and Resources.

K.-F. Fischbach Drosophila Lab - this lab studies brain development in Drosophila melanogaster and is developing 3-D reconstructions of Drosophila brain structures. The lab site includes research reports, publications, a searchable images database, and an extensive links collection. This site is maintained by the University of Freiburg and was an HMS Beagle Featured Lab.