Again, in twenty-one infirmaries with deaths, the average disposable space for each occupied lying-in bed was 2,246 cubic feet; while in nine infirmaries without deaths the space per occupied bed averaged 3,149 cubic feet. These, however, are only averages, and as such may be taken for what they are worth. There were exceptions to these rules in particular cases. The facts regarding Waterford Lying-in Institution have a very important bearing on this question of subdivision. In the years from 1838 to 1844 this hospital consisted of two rooms in a small house. One room was a delivery ward. The other held eight lying-in beds. The total deliveries in this house amounted to 753, and there were 6 deaths = 8 per 1,000. Half this mortality was due to puerperal fever. In October 1844 this hospital occupied another small house, in which the eight lying-in beds were placed in two rooms instead of one as formerly—four beds per room. Up to October 1867 there had been 2,656 deliveries in this house, and 9 deaths—a mortality of 3·4 per 1,000. There were only two puerperal fever deaths in these 2,656 deliveries. These facts appear to show that subdivision among lying-in cases has a certain influence in warding off mortality. But, on the other hand, the death-rates among lying-in cases in particular hospitals are not always in the ratio of the number of occupied beds. A few illustrations of this will suffice. Thus, in the year 1861, there were in the Rotunda Hospital, Dublin, 1,135 deliveries, on which the death-rate was 51·9 per 1,000. In 1828 the deliveries were 2,856, and the death-rate 15 per 1,000. In the four years 1830 to 1833, the deliveries varied from 2,138 to 2,288, and the death-rates were a little more than 5 per 1,000. In Queen Charlotte’s Hospital the highest death-rate occurred in 1849, during which year there were 161 deliveries. The death-rate was 93·2 per 1,000, while in 1832, with 217 deliveries, the death-rate was just one tenth of this amount. In the Maison d’Accouchement at Paris, during the five decennial periods between 1810 and 1859,[9] there were 141,476 deliveries, among which there occurred 6,288 deaths, giving a death-rate of 44·4 per 1,000. The lowest death-rate in any of the decennial periods occurred between 1840 and 1849, when it amounted to 41·9 per 1,000. The largest number of deliveries of any period in the half century was during this ten years. They amounted to 34,776; while, in the period from 1850 to 1859, the deliveries were 24,944, and the death-rate 52 per 1,000. The Dublin Rotunda approximates most to this Paris Maternité in the large number of deliveries, vibrating around 2,000 a year; while, in Queen Charlotte’s Hospital, where, even since its reconstruction, the mortality has been in many years higher than in the Dublin Rotunda, the number of annual deliveries has varied around 200.