Page 142 - Kennemerland VOC ship, 1664 - Published Reports

R. PRICEETAL.:THEKENNEMERLAND
TaMe 2. The Chi-Squaredstatistics measuring the degree to which the distribution of each class of ingot accords with the prediction that they are all equally widely spread
Class
1A
Total Chi-Squared statistic
15 5.93 (n.s.)
IS
2A 15
29
3
4
5A
59
6
7
8
9
10
Unclass. 6
The suffix (n.s.) indicates that the figure computed is not significant statisti- cally, implying no appreciabledeviation from that prediction.
stricken vessel. Otherwise, the Chi-Squaredtest is useful in preventing us from seeking spurious explanations for associations which are super- ficially striking but which are not statistically significant (i.e. they could quite easily have occurred by chance); this includes the clus- tering of class 3 in the middle section of the main gully as well as the similar association of class 7 in the upper part, or the restriction of classes 1B and 10to these two areas.
Apart from the hints of significant concen- trations respecting class 2B and the unclassi- fiable group, the general conclusion must be that overall the ingots are scattered over the site regardless of class, reflecting either the mixing of a stratified cargo during the wrecking, or the mixing of the classes prior to loading. Respec- ting the relationship between sea-bed location and stamps, inspection showed no identity between ingots misplaced in the cluster analy- sis and examples lying apart from the main concentrations of their class. Equally, no sig- nificant variation in ingot weight across the site is apparent when the average for each of the seven discrete areas is calculated.
In conclusion, the main result of this whole analysis has been to contrast the division of this assemblage in terms of shape and stamps with its homogeneity in terms of weight and sea-bed distribution. There are apparently over
a dozen production batches represented, although whether this means casting on different days, by different craftsmen, in different work- shops, or in different countries, we are unable at present to say. The stamps have been shown to be in some way associated with production. But over-riding these divisions there appears to have been a continuing intention to produce ingots of about the same weight, within a fairly wide margin of error. Finally, by the time they arrived on the sea-bed, the several batches had been fairly thoroughly mixed, with a slight in- dication from the significant clustering of two otherwise deviant groups, 2B and the unclassi- fiable ingots, that this mixing took place before or during loading, rather than in the course of the wrecking. Beyond that, pure archaeology can say no more.
Historical observations
In the 17th century the export of lead was a major English industry, the Dutch being the principal customers. The Kennemerland ingots generally conform to the pig-like shape of the English ingots (in contrast, for example, to the ‘salmons’of the French, or the flat rectanguloids of the Spanish) and there is a reasonable claim to consider them of English origin. Each region or area within this country developed its own characteristic forms, markings, and weights, al-
6 3.82 (n.s.) 6.33(n.s.)
5 14.95(s. at 5%)
4 6.24 (n.s.)
3 6.68 (n.s.) 12 3.34(n.s.) 9 2-73(n.s.) 10 4.15 (n.s.) 18 3.18 (n.s.) 2 4.45 (n.s.) 11 2.02 (n.s.)
3
2.38 (n.s.) 18.25 (s. at 1%)
23