"... In PAGE 10: ...resulting in the estimates reported in Table3 . Once again, allowing Student t innovations implies a huge improvement in the log likelihood over a Gaussian specification (see Table 4).... ..."

"... In PAGE 5: ... To be accurate, it is nec- essary to study each case separately, and this has been done mostly during the nineteenth century in the frame of invariant theory. Table1... ..."

"... In PAGE 7: ... To be accurate, it is necessary to study each case separately, and this has been done mostly during the nineteenth century in the frame of invariant theory. Table1 summarizes known values of g(n;; d). Again, these values correspond to the generic case, and there are smaller and larger reachable values (see example in section 4).... ..."

"... In PAGE 7: ...Concurrent with these changes in trade costs and tariffs have been changes in ratio of foreign trade to world GDP, reported in Table2 . In the early nineteenth century trade costs are so high and trade volumes so low (around 1% of GDP) that, of necessity, most production is located close to local markets.... ..."

"... In PAGE 6: ....2. The history of transport costs. While distance remains a barrier even at the start of the twenty first century, the continuing communications revolution has been one of the most outstanding features of the last two hundred years. Table1 reports on the cost of ocean shipping for selected years since 1750. The period between 1830 and 1910 emerges as the era of very substantial decreases and by the late twentieth century ocean shipping rates in real terms were about a sixth of the level of the early nineteenth century.... ..."

"... In PAGE 6: ... A standard indicator is the comparison between trade and GDP growth. As Table1 shows, while merchandise trade has grown faster than output for all periods except between the two world wars, the gap has been larger in the 1990s than in any earlier period since the mid-nineteenth century. More than one- fifth of global output is now exported, double the proportion in the 1950s.... ..."

"... In PAGE 8: ... Over the continent as a whole, the trend of nineteenth-century fertility was no doubt a gentle decline, as compared with a significant drop in mortality; the difference between the two rates of decline accounted for the continent apos;s large natural in- crease. Table2 presents some scattered observations of birth rates and death rates for 1800, 1850 and 1900. In general, the poorer parts of Europe (which were probably also, on the average, the areas of higher fertility and mortality throughout the century) lack data for the early years; there was a rough correlation between prosperity and statistical reporting.... ..."

"... In PAGE 12: ... This is certainly consistent with the predictions of increasing ethnic conflict, although as with the recent decrease of territorial issues in the MID data, one decade of dramatic change can not be considered convincing evidence of a larger trend. [ Table5 about here] Conflict Locations The final set of expectations for future conflict involved the regional distribution of armed conflicts. Table 5 breaks down the three types of COW wars by regional setting, drawing from variables in the COW war data sets that indicate whether or not each war took place within each region; a given war may take place in multiple regions, as with World War II.... In PAGE 12: ... [Table 5 about here] Conflict Locations The final set of expectations for future conflict involved the regional distribution of armed conflicts. Table5 breaks down the three types of COW wars by regional setting, drawing from variables in the COW war data sets that indicate whether or not each war took place within each region; a given war may take place in multiple regions, as with World War II. In the nineteenth century, interstate war was concentrated in the Western Hemisphere and Europe, accounting for ten and fourteen wars respectively; only five nineteenth-century wars occurred in the Middle East and three in Asia.... ..."

"... In PAGE 10: ...40 For name samples comprising between 1,000 and 10,000 names, coding inconsistencies appear to be similar in magnitude to sampling variability. Table1 shows sampling variability for a single name, given different probabilities for the name in the population and different sample sizes. Sampling variability is likely to be insignificant in modern name samples that can easily comprise over a million names.... In PAGE 29: ... 1 C also has some limitations related specifically to name popularity. As Table1 shows, the total popularity of the ten most popular names has changed significantly over the past two hundred years. 1 C measured across a decade in the beginning of the nineteenth century will differ from 1 C measured across a decade at the end of the twentieth century partly because much different weights are attached to the other names category.... ..."

"... In PAGE 9: ...38 For name samples comprising between 1,000 and 10,000 names, coding inconsistencies appear to be similar in magnitude to sampling variability. Table1 shows sampling variability for a single name, given different probabilities for the name in the population and different sample sizes. Sampling variability is likely to be insignificant in modern name samples that can easily comprise over a million names.... In PAGE 28: ... 1 C also has some limitations related specifically to name popularity. As Table1 shows, the total popularity of the ten most popular names has changed significantly over the past two hundred years. 1 C measured across a decade in the beginning of the nineteenth century will differ from 1 C measured across a decade at the end of the twentieth century partly because much different weights are attached to the other names category.... ..."