In 1791 a second attempt to enumerate and classify the population of Cuba was made by order of Don Luis de las Casas. This showed a population of 272,141. This apparently great increase, however, is to be attributed to a more accurate compilation, rather than to any unusual immigration to Cuba during this period. Indeed careful statisticians, notably Baron Humboldt, have reached the conclusion that even these figures fell far below the truth, and that in reality the population of the island at this period numbered at least 362,700 adult persons. Humboldt's conclusions merit quotation. He says: "In 1804 I discussed the census of Don Luis de las Casas with persons who possessed great knowledge of the locality. Examining the proportions of the numbers omitted in the partial comparisons, it seemed to us that the population of the island, in 1791, could not have been less than 362,700 souls. This has been augmented, during the years between 1791 and 1804, by the number of African negroes imported, which, according to the custom-house returns for that period, amounted to 60,393; by the immigration from Europe and St. Domingo (5,000); and by the excess of births over deaths, which, in truth, is indeed small in a country where one-fourth or one-fifth of the entire population is condemned to live in celibacy. The result of these three causes of increase was reckoned to be 60,000, estimating an annual loss of seven per cent, on the newly imported negroes; this gives approximately, for the year 1804, a minimum of 432,080 inhabitants. I estimated this number for the year 1804, to comprise, whites, 234,000, free-colored, 90,000, slaves, 180,000. I estimated the slave population, graduating the production of sugar at 80 to 100 arrobas for each negro on the sugar plantations, and 82 slaves as the mean population of each plantation. There were then, 250 of these. In the seven parishes, Guanajay, Managua, Batabano, Guines, Cano, Bejucal, and Guanabacoa, there were found, by an exact census, 15,130 slaves on 183 sugar plantations." After expatiating on the difficulty of ascertaining with absolute accuracy the ratio of the production of sugar to the number of negroes employed on the different estates, Humboldt continues: "The number of whites can be estimated by the rolls of the militia, of which, in 1804, there were 2,680 disciplined, and 27,000 rural, notwithstanding the great facilities for avoiding the service, and innumerable exemptions granted to lawyers, physicians, apothecaries, notaries, clergy and church servants, schoolmasters, overseers, traders and all who are styled noble."