In estimating the probable accuracy of statistical data in which there may be both excesses and deficiencies, sources of error are diminished by largeness in the numbers employed in striking averages. Bearing this in mind, and after considering the objections brought against the accuracy of the figures, there seems no reason for rejecting the Registrar-General’s average total mortality among lying-in women in England of 5·1 per 1,000, as affording a sufficiently close approximation to the present real death-rate among lying-in women delivered at home, for all practical purposes of comparison with the death-rates in lying-in hospitals. _DEATH-RATES IN LYING-IN INSTITUTIONS._ We shall next show approximately what are the death-rates in establishments for lying-in women. We will give an abstract of mortality statistics for a number of these institutions, the general results of which may be stated as follows:— In eight military lying-in hospitals (Table IV.), in which 5575 deliveries took place, in periods of from 2 to 12 years, there were 50 deaths (excluding a death before admission)—a death-rate of 8·8 per 1,000. TABLE IV.—_Return of the Number of Admissions for Parturition, and Deaths occurring in the undermentioned Women’s Hospitals (Military)._ (_Supplied by the Director-General, Army Medical Department._) ┌───────────┬──────┬──────────┬───────────────────────────────────┐ │ Station │Period│ No. of │ │ │ │ │Deliveries│ Causes of Death │ ├───────────┼──────┼──────────┼────────────┬──────────┬───────────┤ │ „ │ „ │ │ Puerperal│ │ │ │ │ │ │ Feve and│ │ │ │ │ │ │Peritonitis,│ │ │ │ │ │ │ Pyæmia,│ │ │ │ │ │ │ Phlebitis,│ │ Puerperal│ │ │ │ „ │ &c.│Scarlatina│Convulsions│ ├───────────┼──────┼──────────┼────────────┼──────────┼───────────┤ │Devonport │April │ │ │ │ │ │ │1861 │ │ │ │ │ │ │to │ │ │ │ │ │ │Dec. │ │ │ │ │ │ │1869 │ 158│ │ │ 1│ │Colchester │1865 │ │ │ │ │ │ │to │ │ │ │ │ │ │Oct. │ │ │ │ │ │ │187 │ 252│ │ │ │ │Portsmouth │1861 │ │ │ │ │ │ │to │ │ │ │ │ │ │Dec. │ │ │ │ │