NOTE: To see the graphics in this post in focus, just tap your cursor on them.
Throughout the summer months I have been creating Excel spreadsheets containing the skull measurements (for internal cranial capacity) that Morton published during his lifetime. I thought this task would take a few weeks. However, it turned out to be considerably more work than I had anticipated because Morton’s data sets are confusing and not easy to organize. For example, a page of Morton’s 1849 data set is shown below.
It describes one specimen as “16: Iroquois?” and another as “21. Celt: supposed to be a British soldier.” This lack of clarity occurs throughout Morton’s writings.
My goal in creating a modern spreadsheet was to compile a digital data set of Morton’s measurements and then compare them with the measurements I took in 1988. As it happened, that effort was difficult as well. My 1988 data was stored on two 5½ inch floppy disks, but when I searched though my files from 30 years ago, I could only find one of them. I have yet to find a way to retrieve the data on it. Fortunately, I still have all my original hand written lab notes. Unfortunately, there were two possible ways to interpret my notes from back then, and so I am presenting two options for my measurements in Table 1.
Table 1: Morton’s “Shot” Data and Michael’s 1988 Data
In this table “SGM” designates cranial capacity as recorded in Morton’s 1849 catalog.[4,5] “JM3” designates cranial capacity based on the average of all three repeated measurements I took for my 1988 study. “JM2” designates cranial capacity based on the average of just two of the three measurements I took for my 1988 study as explained below.
The reason why I present two measurements for my old data (I will call it Michael88’s data) is quite boring, so I’ll be brief. I measured each skull three times for my 1988 paper. To keep from being accused of having a racial bias, I designed a methodology in which I was “blind” to the results of my measurements until weeks after I measured the skulls. Simply put, I filled the skulls with plastic balls, emptied the balls into a small bucket, and scooped up some of the balls into a small drinking cup. I then weighed the balls and the cup, and repeated the process five times (six times for one big skull). Weeks later I added the weights of all the cups and balls, subtracted the weight of just the cup, and then converted the weight of just the balls to volume. I loaded all the raw data into a computer program, and the program figured how big each skull was. Thus, no one could accuse me of bias in re-measuring the skulls. My original lab notes, which also recorded the words written in the skulls themselves, look like this.
To date, no one has ever accused me of having a racial bias in regards to my research into Morton, so my technique was successful in that respect. However, when I recently put all my data into a spread sheet — and for the first time was no longer blind to it — I found that about eight of my 603 measurements were just dead wrong. They were either far too high or far too low to be correct. Human error had crept into my measurements. Rather than deleting only these dud measurements, I took a different approach. For each skull I measured, I dropped the most extreme value of the three measurements. In other words, I only kept the two measurements that were the most similar to each other (thus generating the JM2 data in Table 1). Indeed, this is not proper scientific technique. And yet, my JM2 data compares quite well with Morton’s, and also with the rigorously scientific measurements reported by Lewis in 2011. I know this because Lewis data posted his 2011 data on the Internet.[6.] I downloaded a copy and compared his measurements with mine.
In 1988 (technically speaking it was 1986), I measured 201 of the skulls that Morton measured using lead shot. The average skull size for Michael88 was 79 in3 while Morton’s mean for those 201 skulls was 81 in3. Of my measurements, 14% (29) of them matched his perfectly, while 4% (9) of my measurements were smaller than his. One quarter (51) of my measurements were within +/- 1 in3 of Morton’s, while 83% (166) were within +/- 3 in3. To put that in perspective, 3 in3 is only about 4% of Morton’s average skull size (81 in3) or Michael88’s average skull size (79 in3). Since I am not a trained statistician, I cannot comment on the statistical validity of the above evaluation. But I can say that over 80% of my reported cranial capacities are quite close to Morton’s. Thus, my findings support Lewis and Michael88 (and also to some extent Gould), all of whom concluded that Morton’s measuring technique using lead shot was indeed sound. Morton’s technique was not perfect, but it was “close enough” for early 19th century science.
My data also compares well with that of Lewis, who measured many of the Morton skulls some decades after I did. Lewis used a more rigorous methodology than I did, and I regard his measurements as the most accurate ever taken. I measured some skulls that he did not, and he measured some that I did not. Altogether, Lewis and I measured 164 of the same skulls. His mean (as I now calculate it) for only these skulls was 77 in3, which was 2 in3 less than mine. Morton’s average for 201 skulls was 2 in3 bigger. Comparing Michael88 to Lewis, a total of 10% (16) of my measurements matched his, while 79% (129) were smaller. A total of 44% (72) of my measurements were within +/- 1 in3 of Lewis’s, while 88% (144) were within +/- 3 in3. To put that in perspective, 3 in3 is only about 4% of Lewis’s average skull size (77 in3). Over 85% of my measurements were quite close to Lewis’s. As previously noted, over 80% of my measurements for 201 skulls were quite close to Morton’s. It appears we all got more or less the same results.
In Stephen Jay Gould’s book, the Mismeasure of Man, he accepted that Morton accurately measured his skull collection using lead shot. However, Morton had initially measured each skull using sifted white pepper seed. The photo below shows Morton’s seed and shot. I found these small spheres lodged into some of the cracks in the Morton skulls when I measured them back in 1986. This photograph presents one of the small seeds (left) and a larger ball of lead shot that Morton used (center), along with an even larger plastic ball (right) I used and a penny to give a sense of scale.
In 1839, Morton published a data set based on his seed measurements. Then in 1849, he published a second data set that was based only on his shot measurements. As it happens, these two data set do not entirely match up in a number of very confusing ways, as I will discuss in a later blog. Nonetheless, I was able to find 35 skulls that Morton measured with seed that I had measured using plastic balls. His mean for this set of skulls was 78 in3 and so was mine. I found that 6% (2) of my measurements matched his, while 46% (16) were smaller. A total of 57% (20) of my measurement were within +/- 1 in3 of Morton’s, while 91% (32) were within +/- 3 in3. And to put that in perspective, 3 in3 is only about 4% of 78 in3, which was both Morton’s and Michael88’s average skull size. Thus, over 90% of my measurements were quite close to Morton’s seed measurements while, as noted above, over 80% of my “shot” skulls were quite close to Morton’s “shot” skulls. Given the small size of this sample, there are no strong statements that one can make about it. But it is the only sample I have, so I present it knowing its limitations.
My conclusion is that both Morton’s early seed data and later shot data are reasonably accurate, and I stress the word “reasonably.” When dealing with old fashioned natural history books and data sets from 150 years ago it is unwise to say that any evidence “clearly” supports any modern conclusion. I have even had problems figuring out my own notes from the 1980s! Nonetheless, like Lewis, I find that Morton’s data sets do not support Gould’s speculative claim that Morton selectively mis-measured the skulls with seed. I am not sure that there is much more that I can add to this statement, other than to stress that I am commenting only on Morton’s ability to measure the internal volume of a human skull. I am not exonerating Morton. His overall research was flawed at many levels, even though his measuring technique generated results with a “reasonable” level of accuracy. And Morton was indeed a racist, although, as I will detail in a future blog, his specific flavor of racism (Anglo-Saxon supremacy) is not the stereotypical racism most people associate with 19th century American historical figures.
But if Morton did a good job of measuring the skulls, is Morton’s overall data set valid? Not really. As I have shown in past blogs, Morton was prone to ignore the information about the skulls that was submitted to him the scholars who sent them to him. For example, Morton received the skull of a Creek Indian from a doctor who had actually dissected the corpse. However, Morton disregarded the doctor’s input. Instead, Morton claimed the Creek was part African based only on the features of the skull. I recently found additional documentation describing how Morton was sent the skulls of Native Americans, but he deemed them to be European or African Americans simply because they did not fit his preconception of what a Native American skull should look like.
While Gould claimed that Morton’s hidden innate racism unconsciously skewed his measuring technique, I propose that Morton’s blatant conscious racism resulted in a pre-sorting of the skulls. I will take up this topic in yet another blog, along with my observations on the confused and disorganized nature of Morton’s data sets. I feel this last issue has not yet been fully explored by Gould, Michael88, or Lewis. As a result, many people (including me) have long been under the impression that Morton’s data sets are clear and easy to interpret, when in many respects they are a mess.
If any readers find some errors in my above data sets, or are otherwise unclear as to what I am saying, please let me know. I had to enter a great deal of data by hand, and my sources were very old books that were hard to read in part due to the less-than-ideal quality of the printing presses used to make them. It is quite possible that I made some mistakes. Furthermore, Morton’s vocabulary is quite antiquated and can be difficult to follow. I have back-checked my work, but another check would be appreciated. I have no panel of professors to review the accuracy of my work, only you folks. Any assistance would be appreciated.
 Samuel G. Morton, Crania Americana: or a Comparative View of the Skulls of the Various Aboriginal Nations of North and South America, (Philadelphia: J. Dobson, 1839), 257-259; Samuel G. Morton, Catalogue of Skulls of Man and the Inferior Animals, (Philadelphia: Merrihew and Thompson, 1849), unnumbered pages 1-63.
 Morton, Catalogue of Skulls, unnumbered page 2.
 John S. Michael, “A New Look at Morton’s Craniological Research.” Current Anthropology 29: 2 (1988), 349-354.
 Morton, Catalogue of Skulls, unnumbered pages 1-63..
 J. Aitken Meigs, Catalogue of Human Crania in the Collection of the Academy of Natural Sciences of Philadelphia, (Philadelphia: Merrihew and Thompson, Printers, 1857), 19-103.
 Lewis, DeGusta, Meyer, Monge , Mann, et al. “The Mismeasure of Science: Stephen Jay Gould versus Samuel George Morton on Skulls and Bias” PLoS Biol, http://www.plosbiology.org/article/info%3Adoi%2F10.1371% 2Fjournal.pbio.1001071, accessed July 2011.
 Stephen J. Gould, “Morton’s ranking of races by cranial capacity: Unconscious manipulation of data may be a scientific norm,” Science, Vol. 200, (1978), 503-509.
 Stephen J. Gould, The Mismeasure of Man, (New York: W.W. Norton, 1981); and Stephen J. Gould, The Mismeasure of Man: The Definitive Refutation of the Argument of the Bell Curve, (New York: W.W. Norton, 2006, 1996).
 Morton, Catalogue of Skulls, vii.
 Morton, Crania Americana, 257-259.
 Morton, Catalogue of Skulls, unnumbered pages 1-63.
 Gould, “Morton’s ranking of races,”506.
 Morton, Catalogue of Skulls,unnumbered page 9.