148 



DISCOVERY 



naturally bclonp. Since these irrcKularitics in no way 

 concern us now, and are indeed disturbing to our 

 present purpose, I have calculated the mean annual 

 mortality from tuberculosis for successive periods of 

 five years, and, having labelled each with the central 

 year of the group, set them in order in the form of 



approximately to the horizontal, and never reaches 

 the base-line at all. For if you halve an apple, take 

 away one part, and halve the remainder, and continue 

 this process, there will always be something left. Of 

 which sort is our curve ? 



The fact is that the rate of the decline of the death- 



Per Million 

 6000 



5000 



3000 



2000 



1000 



Males 



18G1-1870 

 1891-1900 

 1901-1910 



Per Million 



6000r 



4000 



I ! 



5000 



2000 



1000 



I — - 



Females 



1861-1870 

 1891-1900 

 1901-1910 



-i 



rl-^' 



n 



0- 6-10-1&-20-25- 35- 46- 56" 6b- 7o- 0- 5-10-16-20^^25^^^ 36^^ 4 b 65- 



FiG. 1.— THE MEAN ANNUM. DEATH RATE FROM TUBERCULOSIS IN ENGL.^ND AND WAI,ES IN SUCCESSIVI 

 (OVERLAPPING) •■PERIODS" OF FIVE VEAR3. 



Co- ;5- 



the chart which follows, where each column is pro- 

 portional to one of these quinquennial means. 



Even when this has been done, irregularities have 

 not been entirely eliminated, but they have been so 

 substantially reduced that it is comparatively easy to 

 see the general trend of the curve. That curve appears, 

 at first sight, to be a straight line, which, if prolonged, 

 would reach the base somewhere about 1950. A 

 decline following a straight line means, of course, an 

 accelerating rate of dimunition ; for it implies a fall 

 of the same absolute amount each year, and this con- 

 stant decrement must bear an increasing ratio to the 

 declining totals. A constant rate of decline, on the 

 other hand, follows a curve with a convexity down- 

 wards, a curve which is always flattening out, always 



rate from tuberculosis up to 1914 was not a constant 

 but an accelerating rate. Nevertheless the accelera- 

 tion was not such as to produce quite a straight line. 

 The curve does indeed show a tendency, though a small 

 one, to flatten out ; and the complete extinction of 

 tuberculosis may, therefore, be farther off than the 

 chart at first sight seems to indicate. But it must be 

 remembered that tuberculosis would be much easier to 

 deal with were there not so much of it, and that its 

 last remnants should be easily stamped out by segre- 

 gation. We may look forward, therefore, to the problem 

 becoming progressively more manageable as time goes 

 on. After all, then, we need not worry ourselves much 

 as to the exact shape of the curve : the main point 

 is that the extinction of the disease is within sight. 



