276 SEDIMENT OF THE PO. 



It is, indeed, not to be supposed that all the degradation of the 

 mountains is due to the destruction of the forests — that the flanks 

 of every Alpine valley in Central Europe below the snow-hne 

 were once covered with earth and green with woods, but there 

 ai'e not many particular cases in which we can, with certainty, or 

 even with strong probabihty, aflSrm the contrary. 



representative numbers, where a definite is put for an indefinite quantity, A 

 Greek, who wished to express the notion of a great but undetermined number, 

 used "myriad, or ten thousand"; a Roman, "six hundred"; an Oriental, 

 "forty," or, at present, very commonly, "fifteen thousand." Many a tourist 

 has gravely repeated, as an ascertained fact, the vague statement of the Arabs 

 and the monks of Mount Sinai, that the ascent from the convent of St. Cather- 

 ine to the summit of Gebel Moosa counts "fifteen thousand" steps, though the 

 difference of level is only two thousand feet; and the "Forty" Thieves, the 

 "forty" martyr-monks of the convent of El Arbain — not to speak of a similar 

 use of this numeral in more important cases — have often been understood as 

 expressions of a known number, when in fact they mean simply many. The 

 number "fifteen thousand" has found its way to Rome, and De Quincey 

 seriously informs us, on the authority of a lady who had been at much pains 

 to ascertain the exact truth, that, including closets large enough for a bed, the 

 Vatican contains fifteen thousand rooms. Any one who has observed the vast 

 dimensions of most of the apartments of that structure will admit that we 

 make a very small allowance of space when we assign a square rod, sixteen 

 and a half feet square, to each room upon the average. On an acre, there 

 might be one hundred and sixty such rooms, including partition walls ; and, 

 to contain fifteen thousand of them, a building must cover more than nine 

 acres, and be ten stories high, or possess other equivalent dimensions, which, 

 as every traveller knows, many times exceeds the truth. 



The value of a high standard of accuracy in scientific observation can 

 hardly be overrated ; but habits of rigorous exactness will never be formed by 

 an investigator who allows himself to trust im,plicitly to the numerical precision 

 of the results of a few experiments. The wonderful accuracy of geodetic 

 measurements in modern times is, in general, attained by taking the mean of 

 a great number of observations at every station, and this final precision is but 

 the mutual balance and compensation of numerous errors. 



But trustworthy general results can be obtained from means or averages only 

 when the separate items are relatively numerous. The importance of thia 

 principle is well illustrated by the following fact. In an enumeration of the 

 population of London, some years since, conducted with special reference to 

 the effects of different industrial occupations on health, the mortality among 

 manufacturers of children's toys was reported at fifty per cent, per annum. 

 Upon inquiry into that surprising statement, it was found that in London there 

 were two manufacturers of such toys, one of whom had died during the year, 

 and this was the sole foundation for the conclusion that this manufacture was 

 80 destructive to human life. 



The pretended exactness of statistical tables is too often little better than an 



