That Weird Fahrenheit Temperature Scale

(This 736th Buffalo Sunday News column was first published on May 8, 2005.)

We have come to accept some rather strange numbers as part of our culture. Thus, on the Fahrenheit scale which we still use in this country we have 32° as the temperature at which water freezes and 212° the temperature at which it boils. Why those particular numbers? Are they just two more values to make school science tests difficult?

Clearly the Celsius scale with water freezing at 0 and boiling at 100 degrees is easier to remember, but my concern here is not promotion of the metric system. It doesn't need my support. Rather I will explore the history that led to those numbers on the Fahrenheit scale. I was led to this history by my former student, Dipendra Bhattacharya, now a professor at Clarion University of Pennsylvania. He was asked why those numbers were chosen and passed on the question to me. I found the answer in H. Arthur Klein's 1974 book, The World of Measurements, and I summarize his answer here.

Turn back the calendar over 300 years and you will find that there were very few instruments for measuring temperature. Then in about 1700, the Danish astronomer, Ole Roemer, already famous for his proof that light travels at a finite speed, not instantaneously as had been believed, turned his attention to this problem.

Roemer constructed rudimentary thermometers much like those we use today: glass tubes in which alcohol expanded. He was then faced with the problem of how to associate a scale with the rising liquid.

A few years earlier the Italian, Carlo Rinaldi, had proposed that any scale should reflect the freezing and boiling points of water, but Roemer thought that two other temperatures should be taken into account. These were our body heat and the lowest temperature of a mix of water and ice he could attain in his laboratory. He reached this lower temperature by adding salt to the mixture.

So now Roemer had four scale points. He chose 0 to represent that "lowest" temperature and 60 to represent boiling. Why he chose 60 is unclear, but it is a number that appears in minutes and seconds of time and in angle measure. In his rudimentary scale this gave Roemer 7.5 degrees for ice freezing and 22.5 degrees for body temperature.

Enter the German scientist, Daniel Gabriel Fahrenheit. He adapted the idea of Ismael Boulliau to use quicksilver, which we now know as mercury, in his higher quality thermometer tubes. As Klein says, "Had it not been that his thermometers were so sound and practical, the Fahrenheit scale that he devised for them would never have become widely known. As it is, and rather ironically too, Fahrenheit is famous today chiefly because of that same peculiar and highly illogical scale, whereas many of his more substantial contributions to scientific measurement are seldom mentioned."

Taking advantage of the greater accuracy of his mercury thermometers, Fahrenheit decided to expand the Roemer scale by multiplying its values by four. Thus his four scale points were, in degrees: 0 low, 30 ice, 90 body temperature, and 240 boiling.

Those are, of course, far nicer values than those he finally adopted. Unfortunately, he found that they didn't quite serve and he began to tinker with them. First, he found the body temperature too low and he changed it from 90 to 96, as accurate as he could make it but still not the 98.6 degrees we use today. This change, however, raised the freezing point of water to 32 degrees, its present value and the answer to the first part of our problem.

The boiling point too had to be changed. It had not been all that accurately placed on the Roemer scale because the alcohol used in thermometers boils at a lower temperature than water.

So Fahrenheit had to adjust his scale one more time. He reduced his 240 to 212 to fit with his other measures, finally arriving at the three scale points that continue to determine for us Fahrenheit temperatures in degrees: 32 water freezing, our corrected 98.6 body heat and 212 water boiling -- all at sea level, of course. That 0 degrees which provided the scale's starting point is long forgotten.

Klein calls the Fahrenheit scale bizarre and I cannot disagree. One defender, however, notes that for weather it serves quite well, our expected temperatures ranging approximately between 0 and 100 degrees.-- Gerry Rising