Friday, September 3, 2010

The evolution of Cavalli-Sforza. Part I

Germ warfare was a priority for both Allied and Axis scientists. Was one of them a young L.L. Cavalli-Sforza?

With an academic life spanning almost half a century, L.L. Cavalli-Sforza has become perhaps the foremost authority on human population genetics.

He first entered this field in the early 1940s. As a medical student at the University of Pavia, he encountered Emilio Veratti, who lectured on Italian genetics, and Adriano Buzzati-Traverso, who specialized in fruit fly genetics. He went on to study the genetics of bacteria, specifically how to select strains that can better resist radiation and nitrogen mustard (mustard gas):

Starting in 1941, bacteria had become my major interest and in 1948 I gave a paper at the International Congress of Genetics in Stockholm on cross resistance to radiation and nitrogen mustard in E. coli based on work done earlier in Milan with Niccolo Visconti (Cavalli-Sforza 1992).

This account, written forty years after the facts, is consistent with a description he gave in 1950:

Nitrogen mustard resistance was found to be gradual or abrupt in increase in different experiments, only a moderate degree of resistance being acquired, which made a detailed analysis difficult. In E. coli K-12 nitrogen mustard resistance is not accompanied by higher resistance to radiations, as in the case of E. coli B. (Cavalli & Maccacaro 1950)

In the same article, he also described experiments to make E. coli more resistant to chloromycetin, an antibiotic. It is not clear where, when, or for whom this work was done, although some of this research could have been postwar (or early war, i.e., before the U.S. entry in Dec. 1941). The co-author, G.A. Maccacaro, had been supported by a grant from the Rockefeller Foundation, and the K-12 strain had been supplied by an American researcher.

In an earlier article, however, Cavalli-Sforza reported findings from another research project. These findings were the mean death times (i.e., time to death) of mice in response to progressively higher doses of anthrax and pneumococci (Cavalli & Magni 1947). This research was novel in that it measured time to death, as opposed to the percentage killed.

To learn more about this period of Cavalli-Sforza’s life, we can turn to his recent autobiography. Among other things, his research work took him to Germany:

I went to see him [Emilio Veratti] to ask him for advice about certain bacteriological research that I had conducted with Giovanni Magni, my college and faculty classmate. It had to do with a study on bacterial virulence using a mathematical method. Veratti told me that to his knowledge the only person who could help us was a German professor, Richard Prigge. We later discovered he was right.

[…] With Magni, in Como, we studied anthrax, which is still talked about today because it is one of the most fearsome bacteriological weapons

[…] it was not so difficult to measure bacterial virulence [in mice] and we continued our research because we really wished to study this mechanism with a view to creating possible vaccines.

At this point, the advice of Veratti, who had suggested that we try to obtain a scholarship to study at Frankfurt am Main with Prof. Prigge, the only European scientist who could understand the worth of our experiments, proved to be useful. We had the luck of getting a scholarship to spend the summer [of 1942] in Germany.
(Cavalli-Sforza & Cavalli-Sforza 2008, pp. 31-34)

The autobiography mentions another stay in Germany, this time with the renowned geneticist Timofeeff-Ressovsky:

Buzzati asked us, Magni and me, to join him in Berlin, in August 1942, at the Berlin-Buch Kaiser-Wilhelm Institute of Genetics (now the Max-Planck Institute). It was a bit before our research stay in Frankfurt with Prigge.

At the time, the Institute was run by the famous Russian geneticist Nikolai Wladimirovich Timofeeff-Ressovsky, a man of an extraordinary personality, intelligent, likeable, and enthusiastic; in sum, very Russian.

[…] After my meeting with N.W. Timofeef, I decided that I would devote my career to genetics research.
(Cavalli-Sforza & Cavalli-Sforza 2008, pp. 39)

One gets the impression that the meeting with Timoffeef in Berlin happened during a short stay, perhaps a single visit. Yet when Cavalli-Sforza dedicated his genetics textbook to Timoffeef in 1947, he called him “a friend and teacher, with the wishes that he will be able to continue his work” (Cavalli & Buzzati-Traverso 1947). These words suggest a longer working relationship.

Cavalli-Sforza’s autobiography provides the best account of his wartime research, but certain details contradict those in his earlier accounts. For one thing, the earlier ones (with one exception) have him studying E. coli, and not anthrax, although this contradiction may be more apparent than real. He could have been working with E. coli as a prelude to riskier work with anthrax. Alternatively, the E. coli work may have been postwar.

But the contradiction is more fundamental when it comes to his research aims. The earlier accounts have him seeking to create more resistant strains of bacteria, and not better vaccines. These two aims are contradictory because vaccines are normally made from weaker, not stronger strains. Perhaps he also wished to study drug resistance—a legitimate subject of medical enquiry. But why was he also interested in bacterial resistance to radiation and mustard gas?

This sounds more like germ warfare research. More precisely, he seemed to be working on ways to combine anthrax with chemical and radioactive agents, presumably as part of a single warhead.

We’ll probably never know the whole story. Perhaps even Cavalli-Sforza didn’t know. And does it matter? Neither side used germ warfare in WWII, out of fear that the other side would retaliate in kind. (Yes, our side had its own germ warfare program).

Still, it does matter. It certainly did to Cavalli-Sforza. He apparently dreaded having his wartime record brought up, and this dread would guide his behavior later in life ...


During the war years and the immediate postwar era, Cavalli-Sforza published under the name of Cavalli. To date, I have been unable to locate his four wartime articles or even their titles.


Bonezzi, G, L.L. Cavalli, and G. Magni. (1943). Zentralbl Bakteriol, I Orig, 150, 17–25.

Cavalli, L. L., and G. Magni. (1947). Methods of analysing the virulence of bacteria and viruses for genetical purposes, Heredity, 1, 127–132; doi:10.1038/hdy.1947.8

Cavalli, L. L., and G. Magni. (1943). Zentralbl Bakteriol, I Orig, 150, 25–32.

Cavalli, L.L., and G. Magni. (1943). Zentralbl Bakteriol, I Orig, 150, 353–371.

Cavalli, L.L., and G. Magni. (1942). Boll d Soc Med Chir Pavia, 20, 609–624.

Cavalli, L.L. and G.A. Maccacaro (1950). Chloromycetin resistance in E. coli, a case of quantitative inheritance in bacteria, Nature, 4232, 991-992.

Cavalli, L.L. and A. Buzzati-Traverso (1947). Teoria dell'urto ed unità biologiche elementari, Milan, Longanesi.

Cavalli-Sforza (1992). Forty years ago in genetics: The unorthodox mating behavior of bacteria, in J.F. Crow and W.F. Dove (eds). Anecdotal, Historical and Critical Commentaries on Genetics, Genetics Society of America.

Cavalli-Sforza, L.L. and F. Cavalli-Sforza (2008). La génétique des populations : histoire d'une découverte, Odile Jacob.


Tod said...

It strikes me that the early 40's was long before the radiation from nuclear weapons would be thought about in Italy. Improving the vigour and lethality of anthrax by exposing it to stressors such as radiation or nitrogen mustard suggests they were using the stessors to get a stimulatory (hormetic) effect, ending up with a epigeneticly supercharged weapon. In any case it's difficult to see how it was anything but a biological warfare project. In Berlin, in August 1942.

Anonymous said...

Lololozzozl, hormesis is a semi-retarded concept. There is no strong evidence that any common thread runs through the various hormetic phenomena. Some substances simply happen to be hormetic, probably through many different mechanisms, which is not unexpected. Throw a zillion different substances into organisms and you'll find some that are beneficial at certain doses.

Radiation is not hormetic and I greatly doubt that nitrogen mustard is.

I don't like Caballo or Feynman, but germ research is just as necessary as nuke research. If the other bloke might get it, then we better have it also - and we ought to be able to deliver it as a second strike, too. That's MAD logic, and it's about the best you can realistically do in this world. Obviously it would be far better (ceteris paribus) if swords and pikes were the only weapons that could possibly be made, but that is not this world.

Anonymous said...

Well, Feynman's not /that/ bad, so nevermind. I don't really appreciate his writing stuff about promiscuous sex in his autobiographies, but it could be far worse.

Peter Frost said...


Germany had two nuclear weapon projects. One was to build an atomic bomb that would cause destruction through a massive explosion. This project was based at the Kaiser Wilhelm Institute.

The other project (less known) was to build a "dirty" plutonium bomb that would kill through radioactive fallout. It was actually near completion by the end of the war. see:

The second project was probably the one that paid for Cavalli-Sforza's research work. The aim was probably to develop a weapon that could wipe out an army division through a triple combination of anthrax, radioactive fallout, and mustard gas.

Anonymous said...

Oh - you're saying he may have done this for the Axis? Thaaat's a slightly different story, since the Axis had an (objective) deficit in morality vis-a-vis the Allies.

If two sides in a war are on a comparable moral footing I see no problem with any weapons research whatsoever (actually using the weapons as something more than a deterrent is a different matter). Such was the purport of my comment. But the Axis, well...

Peter Frost said...


I agree. If you want to kill lots of people fast, the best way is to use anthrax or nerve gas. A plutonium "dirty" bomb is much more costly and much less effective.

Nonetheless, we know that the U.S. tried to develop a dirty bomb during the cold war. It was a dumb idea, but defence research officers are not always the sharpest tools in the shed.

There is some evidence that Germany had a "dirty bomb" project:

"In 2005, Berlin historian Rainer Karlsch published a book, Hitlers Bombe (in German), which was reported in the press as claiming to provide evidence that Nazi Germany had tested crude nuclear weapons on Rügen island (October 12, 1944) and near Ohrdruf, Thuringia (March 4, 1945), the latter killing 500 prisoners of war under the supervision of the SS. Allegedly, this second test was documented in a film classified after the war in Soviet possession, with the relevant available KGB files also noting highly increased levels of radioactivity in the area of the latter blast. Some press reports, however, have reported the book as only having claimed to provide evidence that the Nazis had been successful with a radiological weapon (a dirty bomb), not a "true" nuclear weapon powered by nuclear fission; while Karlsch asserts that it could have been a thermonuclear hybrid bomb based on the nuclear fusion experiments using shaped charges performed by the Heereswaffenamt under the supervision of Professor Kurt Diebner since 1943."

Anonymous said...

Dr. Frost, could you please enable the captcha word to reduce the spam comments that are filling up many of your comment threads?

Armchair Anthropologist said...

On a different topic, I think that you have made a good case that among Europeans/Caucasians, females have undergone intense selection pressure that has possibly resulted in a lighter skin color and produced a profusion of eye and hair colors.

However, there are two interesting questions that arise.

1. Both males and females are under selection, of course. While at some time in the past there might have been a surplus of females (which can be expected to somewhat relax selection on males, but on the other hand could select females to chose one male as their partner but a more appropriate male as the father of their children, ie, select for skill in cuckoldry, which would then select males for skill in avoiding cuckoldry, via violence or careful selection of females).

2. Females of other racial groups do not show as much variance in appearance as Caucasian groups do. Chinese, or SE Asian, females, for example. Yet, I would have expected those groups to have lived through similar conditions. Are you suggesting that they were hiding out in better climes and rapidly expanded into northern Asian regions after the retreat of the ice?

Peter Frost said...


Can you explain how I can enable captcha word?

Armchair anthropologist,

My argument is that sexual selection of women was strongest in the steppe-tundra belt that stretched across northern Eurasia during the last ice age. On the one hand, steppe-tundra minimizes polygyny because women and children are highly dependent on male mates for food provisioning. On the other hand, steppe-tundra maximizes male mortality because of the long hunting distances involved in tracking reindeer herds (the major food source).

Why then was this sexual selection stronger in the European portion of the Eurasian steppe-tundra? Because the Asian portion was farther north and farther removed from the moderating influence of the Atlantic Ocean. It was much colder and drier. It was polar desert at the height of the last ice age.

There is some debate as to whether the Asian steppe-tundra was continuously inhabited during the last ice age. Most writers say it was inhabited off and on. Sexual selection would have had to restart from zero again and again.

Even if the Asian steppe-tundra had been continuously inhabited, it would have lost much of its genetic variability through repeated population crashes. This may be why hair and eye color polymorphism never developed in northern Asia.

There is a second reason that takes time to explain and most people have trouble understanding it. In High Arctic environments, female infanticide is common because hunting bands are small and far apart. Daughters are killed because they will almost certainly marry out of the band. There is thus no payback for the parents from a future son-in-law. Female infanticide thus 'corrects' for the surplus of excess females and prevents sexual selection from happening.

Anonymous said...

Peter, go to Settings, then click Comments, then click on yes for show word verifications.

Galton's Latent Gemmule said...

CAvALLI, L. & MAGNI, G. (1943). Quantitative Untersuchungen uber die Virulenz. III.
Mitteilung: Analyse der Haufigkeitsverteilung der Absterbezeiten von infizierten Mausen.
Zentbl. Bakt. ParasitKde, Abt. 1, Orig. 150, 353.

"Quantitative investigations into virulence. III. Correspondence: Analysis of the distribution of times to death of infected mice."

Referenced here:

Galton's Latent Gemmule said...

BoNEzzi, G., CAVALLI, L. & MAGNI, G. (1943). Quantitative Untersuchungen uber die
Virulenz. I. Eine Virulenzgleichung und ihre biologische Deutung. Zbl. Bakt. (Abt. 1,
Orig.) 150, 17.

"Quantitative investigations into virulence. I. A virulence comparison and its biological meaning."

CAVALLI, L. & MAGNI, G. (1943). Quantitative Untersuchungen uber die Virulenz. II. Die
bakteriostatische Wirkung der Sulfanilamide in vivo. Zbl. Bakt. (Abt. 1, Orig.) 150, 25.

"Quantitative investigations into virulence. II. The bacteriostatic effect of sulfanilamide in vivo."

Referenced in: