In 1914,
Julian Huxley offered Muller a position at the recently founded
William Marsh Rice Institute; he hurried to complete his
Ph.D. and moved to Houston for the beginning of the 1915-1916 academic year (his degree was issued in 1916). At Rice, Muller taught biology and continued
Drosophila lab work. In 1918, he proposed an explanation for the dramatic discontinuous alterations in
Oenothera larmarckiana that were the basis of
Hugo de Vries's
theory of
mutationism: "balanced lethals" allowed the accumulation of recessive mutations, and rare
crossing over events resulted in the sudden expression of these hidden traits. In other words, de Vries's experiments were explainable by the Mendelian-chromosome theory. Muller's work was increasingly focused on
mutation rate and
lethal mutations. In 1918, Morgan—shorthanded because many of his students and assistants were drafted for the U.S. entry into
World War I—convinced Muller to return to Columbia to teach and to expand his experimental program.
At Columbia, Muller and his collaborator and longtime friend
Edgar Altenburg continued the investigation of lethal mutations. The primary method for detecting such mutations was to measure the sex ratios of the offspring of female flies. They predicted the ratio would vary from 1:1 due to recessive mutations on the X chromosome, which would only be expressed in males (who lacked the functional allele on a second X chromosome). Muller found a strong temperature dependence in mutation rate, leading him to believe that spontaneous mutation was the dominant mode (and to initially discount the role of external factors such as ionizing radiation or chemical agents). In 1920, Muller and Altenburg coauthored a seminal paper in
Genetics on "modifier genes" that determine the size of mutant
Drosophila wings. In 1919, Muller made the important discovery of a mutant (later found to be a
chromosomal inversion) that appeared to suppress crossing-over, which opened up new avenues in mutation rate studies. However, his appointment at Columbia was not continued; he accepted an offer from the
University of Texas and left Columbia after the summer of 1920.
Muller taught at
The University of Texas from 1920 until 1932. Soon after returning to Texas, he married mathematics professor Jesse Marie Jacobs, whom he had courted previously. In his early years at Texas, Muller's
Drosophila work was slow going; the data from his mutation rate studies were difficult to interpret. In 1923, he began using
radium and
X-rays, but the relationship between radiation and mutation was difficult to measure because such radiation also sterilized the flies. In this period, he became more involved with eugenics and human genetics. He carried out a study of twins separated at birth that seemed to indicate a strong hereditary component to
I.Q. Muller was critical of the new directions of the eugenics movement (such as anti-immigration), but was hopeful about the prospects for positive eugenics.