Archival research can sometimes be very surprising. I experienced this firsthand last week when I doing some research for a manuscript that I am coauthoring with a colleague.

I was reading some historical scientific articles when I stumbled upon a book chapter entitled “A Study of a Pair of Siamese Twins” (Koch, 1928). Of course, something like that grabs anyone’s attention.

The study reports information about a pair of female conjoined twins called “D” and “V.” They were born in Brighton, England, and were 14 years and 10 months old when tested at “Christmas, 1922,” which would mean they were born about February 1908. The twins had spent most of their life “associated with travelling shows or vaudeville circuits” (Koch, 1928, p. 77).

This sounded very familiar. Years ago I had watched a documentary on Netflix about Daisy and Violet Hilton, conjoined twins whose life fit this description exactly. I also knew about them from the stage bio-musical Side Show, a play no one needs to see. (They also were in the cast of the 1932 film Freaks and the 1951 film Chained for Life, though I have not seen either one.) Could this be a report about one of the most famous sets of conjoined twins in history?

It was possible. At the time, psychologists made little or no effort to conceal subjects’ identities (Powell et al., 2014). Participants in studies were often listed under their real name (e.g., Terman & Fenton, 1921), and subjects were often reported under their initials or with names similar to their real ones. For example, one of the candidates for John Watson’s famous “Little Albert” was named Albert Barger (Powell et al., 2014).

When checked against the twins’ Wikipedia page, more details aligned with Koch’s (1927, 1928) reports:

  • D is the “right-hand member” (i.e., the twin on the right from their perspective), which was true of Daisy.
  • Their career in show business took them to Europe, Australia, and North America.
  • The two lacked formal schooling, but were proficient musicians.
  • They were accomplished dancers.

There were some slight differences between the two descriptions. While both sources say that they were proficient musicians, Koch (1928) says that they could play the saxophone, clarinet, and piano, while Wikipedia says that Violet played the saxophone and Daisy played the violin. Another discrepancy is Koch’s (1928, p. 77) claim that they were “orphans from birth,” whereas Wikipedia says that their unwed mother gave them away during infancy to a woman who prepared them for a career in entertainment.

Although not all the details match perfectly, I thought it was extremely likely that these report were a case study of Daisy and Violet Hilton. How many female conjoined twins born in February 1908 in Brighton, England, and living in the United States on the vaudeville circuit could there have been in 1922?

My suspicions were fully confirmed when I compared the photograph published in Koch’s (1927) report with a picture of Daisy and Violet Hilton taken in the early 1920s. These are clearly the same people. Daisy, on the viewer’s left, is even smiling the exact same way in both photographs and has her head tilted at almost the same angle. Even the hairstyles are the same.

Left: Photo of “D” and “V” conjoined twins from Koch (1927, p. 316). Right: Photo taken in the mid-1920s of Daisy and Violet Hilton (source of right photo).

The twins took three full intelligence tests: the Terman Group Test of Mental Ability, the Army Alpha, and the Army Beta. On all three tests, these two examinees scored lower than average when compared to other samples of adolescents. On the Army Alpha, their IQ scores were approximately 63 (for Daisy) and 58 (for Violet). On the Army Beta, their scores were higher: about 86 (for Daisy) and 79 (for Violet). Their IQ scores on the Terman Group Test were approximately 81 (for Daisy) and 73 (for Violet). They also took a several additional subtests, including digit span (earning scores of 5 for Daisy and 6 for Violet), and nonverbal subtests, such as an object completion task called the Manikin Test.

The low IQ scores are unsurprising because the Hilton twins never had formal schooling, and often the two were abused and neglected during their childhood. The fact that they performed best on the non-verbal Army Beta test–which has no content that resembles school tasks–I think is telling. Moreover, Koch’s (1927, 1928) reports indicate no hint of intellectual disability. In fact, Koch (1927, p. 319) stated that both twins enjoyed reading detective stories–an uncommon past-time among low-intelligence populations. If this case study reveals anything of scientific interest, it is probably the importance of schooling for helping people solve intelligence test tasks.

Finally, there were several academic tests that the twins took. They scored very poorly on these compared to other samples of adolescents. As an example, Daisy apparently had not mastered multiplication yet, and Violet only got one multiplication problem correct out of three. A comparison sample of eighth-graders attempted more questions (10, on average) and got more right (an average of 5). Again, for examinees who had never had formal schooling and whose “education” was focused on performing, these results are not surprising.

Koch’s (1927, 1928) reports include data on many other variables, including physical measurements. (The two were very short for their age: Daisy was 110.7 cm tall, and Violet was 106.7 cm tall. The average at the time was 150.84 cm.) The two also took a word association test and some early personality tests (the Downey Will-Temperament test and Pressey X-O Test of Emotional Attitudes). Personality assessment has matured so much in the past 98 years that I do not think that these scores say anything of scientific value, but a biographer would probably be interested.

In my Google searches, the only people who seems to have publicly made the connection between Koch’s (1927, 1928) reports and the Hilton twins are Cummins and Mairs (1934) in an obscure article about fingerprint similarity among conjoined twins. Koch’s reports seem to have made little impact. The 1927 article has only been cited 14 times, and the 1928 chapter has only been cited twice (both citation counts from Google Scholar).

So, I seem to have uncovered a some biographical details that have escaped the notice of historians thus far. It pays to spend time reading older articles. You never know what you will discover.


Cummins, H., & Mairs, G. T. (1934). Finger prints of conjoined twins: Differences in Gibb twins and their genetic significance. Journal of Heredity, 25(6), 237-243.

Koch, H. L. (1927). Some measurements of a pair of Siamese twins. Journal of Comparative Psychology, 7(4), 313-333.

Koch, H. L. (1928). A study of a pair of Siamese twins. In G. M. Whipple (Ed.), The twenty-seventh yearbook of the National Society for the Study of Education: Nature and nurture: Part I, their influence upon intelligence (pp. 75-80). Public School Publishing Company.

Powell, R. A., Digdon, N., Harris, B., & Smithson, C. (2014). Correcting the record on Watson, Rayner, and Little Albert: Albert Barger as “Psychology’s lost boy.” American Psychologist, 69(6), 600-611.

Terman, L. M., & Fenton, J. C. (1921). Preliminary report on a gifted juvenile author. Journal of Applied Psychology, 5(2), 163-178.