ROBERT PAGE, Feb 12th 2015, fig.5 updated Nov.23rd 2015.
Fig.9. Combined data for Nicole and Henri at London, Nicole at Canterbury, Ion at Bury.
As the dies were all cut by the same group of people in London, it is valid to combine the above four data sets to better see the big picture. The resultant graph appears below, but keep in mind that all classes are not represented at all the mints.
Reverses from Bury come in two main forms ION/ONS/EIN/TED or ION/ONS/EDM/VND – however both types have three quadrants with the letter N – represented of course by either an N or an H.
The sample sizes are smaller at Bury, so the results for classes where there are less than 20-30 samples are not as reliable – classes with less than 20 coins are 2a, 2b1 and 2b2, 3ab2 and 4ab. Samples for the three Class 2 sub-classes have been combined. Looking at class 2 as a whole we see all three types are present, but that HHH disappears after 3ab1 and makes a minor comeback with the last coins of Ion in 5a3. 3a2 appears anomalous, and the 3a2 pattern would fit in better between 3c and 3d1 – as seen for Henri of London, but not Nicole. The same predominance of NNN is seen for the second half of class 3, except for the re-emergence of a significant fraction of mixed combinations for 3d1.
Fig.8 - Ion at Bury St Edmunds
Having looked at London and Canterbury, where there are a statistically robust number of samples for most classes, I turned my attention to one of the lesser mints to see if the observations would hold true for a mint with a lesser number of coins. I examined the statistics for the coins viewed by Eaglen from the Bury mint, as documented in Churchill and Thomas (2012). The resultant graph is as follows:
Fig.7 Nicole of Canterbury
Observations from Both Nicole and Henri of London
Henri did not start issuing coins until class 3 so we only have the evidence from Nicole for the letter usage for class 2, and we see that the HHH combination was prevalent then, with it reducing significantly from 3a onwards. The fact that the use of HHH reduces from class 2 through class 3 and is eliminated by class 4 suggests that HHH was recognised as a less correct way to letter the coins, resulting in not only the elimination of HHH but also mixed combinations by class 4. The number of class 4 samples is limited, and one may yet turn up with a legend other than NNN. When the attention to detail dropped of after classes 4 and 5a we see usage of mixed combinations, and even the occasional use of HHH coming back.
To test the observations from London, I have below plotted the data for Nicole of Canterbury...
Observations are as follows:
Fig 6 Henri of London
Having looked at Nicole, I turned my attention to Henri, to see if the same observations hold true.
Henri: Variations through Time
A similar graph was constructed for the Henri of London coins, but with a few more classes shown.
Observations are as follows:
Fig 5 : Relative Proportions of NNN, HHH and Mixed Combinations
Variations Through Time:
One can look at the relative proportions of combinations in the above examples, and from other classes with a large enough sample in order to determine any trends through time. For Nicole, I have examined the data from Class 2 through to class 5c2, omitting some classes with a small sample size. The last issue by Nicole was 5c2, and the last coins issued under his control may have been issued when he was ailing and perhaps not in a position to pay as close attention to supervision as he had in the past.
The resultant graph for Nicole at London is shown on the right.
Fig.4 Nicole of London, class 5b2
In this class 5 example on the left, NNN clearly again predominates, followed by various mixed combinations. Of the 1809 coins in the sample, only three have the HHH combination; so these are quite rare.
Fig 3 – Nicole of London, class 3b
Again, in class 3b, we see NNN as dominant, with HHH being less common than the sum of all the mixed combinations. In fact, HHH makes up only about 7% of the total coins, so is quite a scarce 3b variety.
Fig.2 Nicole of London, Class 3a2
In this next figure, on the left we see a very different pattern; the HHH combination is now far less common than NNN, and four mixed combinations are seen, each scarcer than HHH.
Fig.1 Nicole of London, Class 2
In this first figure on the right we see the H/N usage for class 2 coins. Clearly HHH predominates and anything else is relatively scarce. Of the various combinations HHN (e.g. HIC/OLE/OHL/VND) is the least scarce. The all N combination, NNN, does exist but is very scarce.
Summary: A statistical analysis of the reverse legend on some of the coins from the Brussels Hoard reveals changes in the usage of “H” for “N” through time. The pattern of usage suggests that “N” was seen as a more correct style than using an “H” for an “N”. Although the variation through time is now better documented, the reason for the existence of coins with a mix of “H” and “N” in their reverse legend is not yet understood. The study also highlights some rarer letter combinations for collectors to seek out.
Throughout the Henry III long cross series one often sees the capital letter “H” often used for an “N”. This usage is not consistent, in that many coins have a mixture of “H” and “N” in their reverse legends. I have looked at the two moneyers, Nicole and Henri, as both have an N in their name and both moneyers were active over much of the long cross coinage period.
The combination of moneyer and London mint name gives 3 examples of letter N/H to examine – normally in the 1st, 3rd and 4th quadrants. For example NIC/OLE/ONL/VND I record as NNN, and HIC/OLE/OHN/VHD as HHH. Mixtures of “H” and “N” on the same coin are recorded as MIX, and on some graphs I have broken out the individual mixed combination.
The data shown here is based on coins in the Brussels Hoard as documented by Churchill and Thomas (2012). Unfortunately the catalogue of the Colchester Hoard coins doesn’t distinguish between the use of H or N.
Firstly I will show some individual class examples for Nicole, after which I will look at the variation through the series for Nicole, and then do the same for Henri. Nicole coins of Canterbury and Ion of Bury have also been analysed. Examination of the plots allows various conclusions to be drawn.
It is hard to understand why a mixture of H and N should be used on the same coin, if all the legend was completed by the same die cutter. On coins with three letter H/Ns on the reverse there are eight possible combinations – could these different combinations have perhaps been signatures of different die cutters? This is doubtful given the wide variation in the proportion of mixed to non-mixed combinations. The reasoning behind the mixed usage still remains a mystery.
One aspect of the letter N is that sometimes there is a pellet on the cross-bar of the letter. These pellets appear to be random, but may have some significance. This occurrence of pellets on the letter N will probably be the subject of a future research note.
However this small study has served to indicate some of the scarcer lettering combinations in each class that collectors may be interested in seeking out.
Looking at both the individual data sets as well as the combined data set above, we can deduce that NNN was regarded as the correct usage, but that mixtures of H and N were considered acceptable for much of the long cross period, except during the brief class 4 period when we see the highest quality of die cutting.
At the beginning of Class 2 of the voided long cross series it seems that HHH was the widely accepted usage. This changed abruptly at the beginning of class 3, with the fraction of NNN coins generally increasing through class 3, with 3ab2 being a notable exception. This abrupt change of HHH usage at the class 2/class 3 transition may have been due to an influx of new die cutters to cope with the needed production volume. On fig.9 the pattern for class 3a2 seems to fit correctly in sequence, though the high proportion of mixed combinations of class 3ab2 is anomalous.
The HHH coins had almost disappeared by class 5. Though a scarce few are seen. The fraction of mixed H/N coins is generally higher for class 5 than class 3, perhaps indicating a relaxation of the standards that had been achieved by Class 4 time, possibly related to some of the die-cutters needed for the provincial phase being "let go".