The spirit of improvement thus awakened exerted itself with increased effect during the whole of the eighteenth century. Agriculture, which was now rapidly advancing, had created a demand for town refuse, the fertilizing property of which began to be perceived; so that all manner of offensive substances were regularly carried away to the fields, to the great increase of the cleanliness of the streets. At the same time many of the narrower streets were widened, the houses were entirely taken down and rebuilt, and in this operation slate was universally substituted for thatch, and brick for timber. The pavement also, which had long been the reproach of London, was improved. Population in the mean time rapidly increased, less by the relative increase of the number of births than by the proportionate decrease of the deaths, and this notwithstanding the occasional occurrence of severe pestilence. The result of the whole was an increase in the length of life. An increase in the length of life is an expression and a measure of the sum of comfort experienced from the whole collective circumstances that make up national prosperity. In the interval between the seventeenth and eighteenth centuries that sum grew into a highly important one. Of this the proof is positive. It happened that in the year 1693 a loan was raised for the service of the State by the method of Tontine, and that another was contracted by the same method in the year 1790; the interval being almost exactly a century. The term Tontine is derived from the name of the originator of this scheme of life annuity, the principle of which is this. The person who advances £100 is at liberty to name any life he pleases, during the existence of which he draws a certain annuity; and as the shares of the dead nominees are distributed among the living ones, the annuity continually increases till the last survivor gets the whole income. A comparison of the experience between two Tontines gives the exact measure of the effect produced on the duration of life, by such changes in the social condition of the people as may have occurred in the interval between them. A person of the male sex (for there is a considerable difference in the results in the two sexes), living in 1793, compared with a male living in 1690, at fifteen years of age, had gained an expectation of life of nearly ten years; at twenty years of age, nine years and a half; at twenty-five years of age, upwards of eight years; at thirty years of age, upwards of seven years, and so on. Or the gain in the expectation of life may be stated more correctly in years, thus: Take for example a man at the age of 30, in 1693 his expectation of life would have been 26.665; in 1790 it would have been 33.775 years. On this evidence Mr Finlaison justly observes that civilization could not have increased by a single leap in the time of Mr Pitt, but must have been slowly on the increase at least since the days of Queen Anne.