Clinton's Serial Numbers
Serial numbers are normally serial - that is to say they follow upon each other without gaps. But if this is the case with Clinton's serial numbers, it points to a very high productivity (about 7.5 flutes per working week) for a small and relatively short-lived company , and a high sales rate at a time when competition between all the new models on offer was absolutely intense.
Given the seemingly large number of flutes made in a very short time, it is possible that the serial numbers are not contiguous - for example that series of numbers might have been used for different kinds of flutes. If so, we have so far been unable to crack the code. The Flutes for India suggest that they might be in a series starting at 2000, but the closely spaced group starting at 4803 contains a wide range of flute types. It is of course possible that he ran with one philosophy of serial numbering for a while then dropped it in favour of some other system. Or indeed that the system broke down after Clinton's death.
The alternative suggestion is that Clinton simply fudged the figures to make sales look more impressive. A fair comment is where are all these 6000 flutes today? You might imagine then that Clinton might have left a few spaces between each number issued, possibly even varying the size of these gaps to avoid being outed. We'd expect then to see the numbers on extant flutes to be well distributed between 1 and 6000 or so. But when we look at the distribution of serial numbers of extent Clinton flutes we see a very different pattern.
The extant flutes fall into clear groups where:
This distribution with its sudden big jumps doesn't seem to fit well with the fudging suggestion, yet neither can we see logic in the groupings of flute types. Perhaps the most peculiar is the sudden jump from around 2000 to near 5000, with so far only 4 flutes found in the gap.
The system might of course relate to something of which we have no knowledge - changes in personnel, workshop practice or whatever, that might have been significant to Clinton but to nobody else. Or indeed, we could be reading too much into the small amount of data we have. Hopefully, as we become aware of more extant Clinton flutes, the picture will clarify.
Given our doubts then that the numbers are contiguous, can we estimate a more likely output from the Clinton workshop over the 16 years? Perhaps one approach would be to separate the groups of numbers with the average spacing within the groups, 25. This would suggest a figure of 903 flutes, equivalent to an average rate of manufacture of just over 1 a week. This figure is more in keeping with what we might expect of a small manufacturer like Siccama or Hudson and indeed quite good going, given Clinton's flutes are generally more complex.
Indeed, this might be too conservative a figure, especially as we currently aware of nearly three times as many extant Clinton's as extent Siccama's. Looking again at the chart above, we wouldn't need to find many more flutes to close up some of those early gaps, quickly adding several hundred to the output. But this is becoming speculation upon speculation, so perhaps we'd best wait to see what turns up!