second. But the portion of a roof drained by each pipe being a very small fraction of an acre, the quantity of water flowing through a roof pipe per second seldom exceeds a fraction of a gallon, while roof pipes are rarely less and often more than three inches diameter; and if flowing full at the rate of thirty feet a second (which is very moderate for a vertical pipe from a roof) would discharge at the rate of nine gallons per second. How then is it possible that, if unobstructed by leaves or other solid matter, they can ever be so filled as to prevent air escaping up them? If these objectors had even made the most cursory observation of what really occurs during rain, they never would have imagined such an absurdity; but it is easier to trust to the imagination than to observation for facts; but no one ever saw a stream of water flowing from a roof gutter fast enough to fill a roof pipe a quarter its size. That streams of greater volume do not run into them is proved by the very common practice of conveying the water from the gutter into the roof pipe through a short length of pipe of about half its diameter, or one quarter its area, which smaller pipe if unobstructed never runs full, though the rate of flow is much slower than it would be in a much longer vertical pipe if that were running full. Those therefore who have urged and repeated the objection that roof pipes get filled with water by heavy rain, prove not only their own want of observation but also their ignorance of or inattention to the most elementary facts of hydraulics and rainfall. If nearly all the roof pipes in a town were used to provide for the escape of air from the sewers (which, as it is generally warmer than external air, would generally rise in them) so little impurity would escape by any that it would rarely be annoying or dangerous, and, when it was so, annoyance could easily be guarded against by using Dr. Stenhouse's air-filters. Such filters, however, to be trusted, must be constructed as their inventor proposed, not in bad imitation of them—that is, the air to be purified must go through, not merely pass over, the purifying charcoal, to secure the necessary absorption and decomposition of any offensive gas, and, what many consider more vitally important, to prevent the escape of any zymotic germs the sewer air may contain which may not be detained by merely passing the air over and not through a layer of charcoal; the mere fact of the air being rendered. inoffensive being no proof that it is made safe. It has been objected that a layer of charcoal through which the air must pass will act as a plug and check the current; but if the air filter be large enough it will check the current very slightly; it will not act like a cork, and will not confine the air in the

sewer and compel it to force its way into the houses through the traps, nor will it ever prevent a considerable quantity of air passing quite as fast as it should pass, if its complete purification is to be depended on; faster than that is neither necessary nor desirable, but it is desirable that a considerable quantity of charcoal should be used so that it will not need to be renewed for a long time. If many of the roof pipes be used for this purpose, and nearly all may be, the quantity of air passing by each will be very small, and as that small quantity can easily pass through a real air-filter there is no excuse for resorting to a sham one.

1

The Vital Statistics of the Coastal Districts of England and Wales. By ALFRED HAVILAND, Medical Officer of Health to the Combined Sanitary Authorities in the counties of Northampton, Leicester, Rutland, and Bucks.

NGLAND being a triangular peninsula, may be divided as

regards its coastal districts into three lines,-the eastern from Berwick to Romney Marsh, the southern from Rye to Penzance, and the western from Redruth to Carlisle. The eastern line from Berwick to Flamborough Head, with the exception of the mouth of the Tees Valley, presents a somewhat bold coast, which has a protective influence on the sub-coastal and inland districts. From this point to Romney Marsh the average height of the cliff is low, and admits freely the prevailing winds over a large extent of country. The average annual death-rate of this line, consisting of fifty-four coastal districts, is 20.25, or 2.17 per 1,000 less than the average for England. It shows, however, when compared with the standard rural group, a death-rate nearly four per 1,000 higher (3.85). It must be borne in mind that the high mortality districts on the Tyne and Humber influence to a certain extent not only the general death-rate, but the causes which conduce to their high mortality seem to extend on each side of them. The normal death-rate of this coast under present conditions would belong to the fifth degree, or equal to a mortality per annum of from 17 to 19 per 1,000.

If we now look at the map of mortality from all causes at ages under five years, we find, perhaps, a general resemblance to that at all ages, although it is evident from the greater extent of blue districts along the Northumberland, Durham,

1 See Sessional Proceedings, 1871–2, p. 203.

and Yorkshire coasts, and the greater number of districts covered pale pink, indicating the fourth degree, that the mortality is proportionably greater at this tender age by the seacoast than during riper years. Among children under five years the death-rate along this coast is 56.06 per 1,000, or 15.90 per 1,000 more than in the standard group of rural districts, wherein it equals 40.16 per 1,000; and compared with the mean of the whole coast-line of England it shows a death-rate above the mean equal to 2·12 per 1,000. The normal mortality would under present circumstances be that of the fourth degree, or only one degree below the average. This is a high and significant death-rate, especially when we consider the areal extent of the coast group.

On carefully viewing the coastal districts on the map, showing the death-rate from all causes at all ages, from Berwick to Carlisle, it will be found that one alone is characterised by the much to be desired sixth degree; this is Steyning, at the south-east corner of which lies the old town of St. Brighthelmstone, enjoying a death-rate of 22.02 per 1,000, or 0·40 below that of England and Wales. Steyning belongs to a group of low mortality districts, and is its coastal representative; the others, East Grinstead, Cuckfield, and Thakeham, lie inland. This group has a mean death-rate 16.16, or 6.26 per 1,000 less than that of the country in general, and 0.34 less than the standard group of rural districts, the death-rate of which amounted to 16.50. This is one out of the many facts that these maps show, and prove that the coastal districts cannot boast of superior healthiness to those lying inland.

Let us take, for instance, another group inland-the group, in fact, which is taken as the standard for the inland rural districts.

Herein we find the annual death-rate to equal 16-50 only. The constituent districts are Epsom, Hambledon, Dorking, Reigate, Godstone, and Bromley. The lowest in this group is Godstone, 15.56. In the Sussex group it was Cuckfield, 1571, so that even here the more inland district shows the lower death-rate. If we now examine this coast on the diarrhoea map we shall see the sixth or lowest degree of mortality appears for the first time. This degree indicates a death-rate of from one to two in every 1,000 children living under five years of age. We find these districts arranged into four groups 1. The Northumberland; 2. Lincoln; 3. Norfolk; and 4. Suffolk. These groups have certain features in common. None of them are near the great high mortality centres, and they all occupy prominent coastal positions; three out of the

four groups being so situated as to receive the full influence. of the sea winds.

If we proceed along the south coast we find again these low mortality diarrhoea districts, grouping themselves:-1. The Kent group, including Romney Marsh, Rye, and Battle, the two former well exposed to the south-easterly winds; 2. The Dorset and Hampshire group, which is intersected by the Southampton Water and natural harbour of Poole; 3. The Devonian group; and, lastly, that of Cornwall.

The average death-rate from this cause is 3.66, or nearly one per 1,000 less (viz., 1.71). The normal death-rate from diarrhoea under five years along the coast belongs to the fifth degree, or the last but one, comprising from 3 to 4 deaths annually per 1,000. Along this coast we have no deathrate of the first degree. Brighton and Plymouth may be bracketed together as belonging to the second degree, 9 to 10 per 1,000.

The low mortality groups of the sixth degree have all, with one exception, prominent headlands, viz., Dungeness, Selsey Bill, St. Alban's Head, Start Point, and Lizard Head. Let us now round the Land's End, and proceed along the coast of Cornwall, Wales, Lancashire, and Cumberland. In the map of Deaths at All Ages, and in that of Deaths from All Causes under five, we find no degree of the sixth power, but we discover the dreaded blue mark of the first, limited, however, in both instances to Liverpool.

On this coast we all know that the south-westerly, the westerly, and the north-westerly winds, all of which are sea winds, blow with full power. These winds are our prevailing winds, and bring with them from the equator plenty of oxidising ozone-too much for delicate lungs; they still bring health to others who suffer from the many causes producing diarrhoea, amongst the most prominent of which are sewage

gases.

This

On this coast we find the sixth degree prevailing, giving a marked feature to the coast, especially along that of Wales, every district of which, except Cardiff, is coloured deep red. is in accordance with what has been seen on the heart disease map. Here all the valleys are in the direct axis of the prevailing winds. They get well air-flushed by the most powerful of our sea winds, and coincident with this we find the least amount of mortality among children from diarrhoea.

This coast has a death-rate identical with that of the standard group of rural districts. viz., 2.85. The group of high mortality includes Liverpool and Preston, which have death

rates from this cause equally, respectively 10:00 and 15·00 per 1,000, or 7·15 and 12.15 in excess of that which obtains in the rural districts.

Before proceeding to another map I will just summarise what I have already stated.

1. The eastern coast has a higher general mortality from all causes at all ages than the standard rural districts group, as 20.25 is to 16.50.

2. That under five years of age the mortality is also greater, as 56.06 is to 40.16.

3. When compared with the death-rate of England and Wales it is as 20.25 to 22.42 at all ages, and 56′06 to 68.30.

4. From diarrhoea, under five years, the death-rate was at 4.68 on the east coast to 2.85 in the rural group, and 5·98 in all England and Wales.

5. The south coast had a mean mortality at all ages of 19.78, that of the rural group being 16.50. Under five years the death-rate was 52.14 along the south coast, and 40·16 in the rural group. From diarrhoea the death-rate is 3.66 on the south coast, 2.85 in the rural group, and 5.98 in all England and Wales.

6. The western coast had a general death-rate at all ages of 20.72, under five years 53 62, and from diarrhoea 2.85, as compared with 22.42, 68.30, and 5·98. All these figures undoubtedly prove that, generally speaking, there is a less amount of mortality, and therefore a greater amount of health, to be found along the coast than there is generally throughout the country; but when we compare the coastal death-rate with inland, we are at once struck with the fact that the former is not to be compared with the latter. Only one district of the sixth degree is to be found in the maps of general mortality among all the 147 coastal districts.

This teaches us a lesson which ought always to be remembered. The death-rate throughout our country is under our own control to an enormous extent, and it matters not how favourably we may be situated as regards our insular position, receiving as we do the full effects of our prevailing sea winds, if we neglect the principles of hygiene. This high coastal death-rate need not exist, and I have selected deaths from diarrhoea under five years of age, as I believe it to be one of the best exponents of insanitary conditions that we can find. Show me a high death-rate from this cause, and a tale is instantly told of overcrowding, sewer gases, bad drainage, bad water, bad food, and unhealthy mothers. If we would earnestly wish to make our watering-places what really they ought to be-health