Then, again, in the consideration of the foreign duties which devolve on our forces, it is often urged that long service, at all events, is best adapted to meet Indian requirements, on the two grounds that young soldiers cannot stand tropical climates and that frequent reliefs are costly. Neither of these views will bear the test of examination. The report of the sanitary condition of the army in India said that 'upon the whole, early entry into India appears to be an advantage, not only at first, but in after life.'[88] At a subsequent period, Sir Ranald Martin stated that 'all statistical observations go to disprove anything like acclimatisation in the East Indies.' On the contrary, he declares that 'disease and death increase with length of service and age.' Dr. Brydon also said: 'The death-rate of 1871 shows that the death-rate for the men above thirty has been consistently double that of men below that age.' Lord Airey's Commission of 1880[89] quoted figures proving that the proportion of deaths, and of invalids sent home, increases in a rapid ratio with age. For instance, the number of deaths and invalids per 1,000 men on the average of ten years is-- Deaths Invalids Under 25 years old 16·06 25·84 Over 35 and under 40 33·71 76·11 The above figures indicate that men should be sent to India young, and not be kept there beyond a few years. If men, whose service in India is prolonged, die twice as fast after thirty as they do when under that age, it is evident that, as they must be replaced by drafts from home, even from a financial point of view such a system is not to be commended. So far from a short period not being adapted for India, it is the only one which ought to be allowed, on grounds alike of humanity, efficiency, and economy. There still remain a few other points, regarding the men serving in the present day, which are worthy of notice. The effects of the Education Act of 1870 have been very marked on the Army. For instance, the proportion of men in the ranks of what is termed 'superior education' was in 1861 74 per thousand, whereas in 1889 it had risen to 854.[90] Again, as regards crime, in 1868[91] the proportion of courts-martial per thousand was 144, whereas in 1892 it was only 54.