Solution B.-Sodium nitrite (not nitrate), per cent. (of I per cent.) solution in distilled water.

Solution C (Test Solution).—One hundred parts of A plus one part of B.

The original test solution of Ehrlich was made by adding to 40 parts of solution A 1 part of solution B. This the author has modified by using 100 parts of A to 1 of B, with the effect of eliminating many disturbing factors and doubtful reactions. Dr. C. E. Simon suggested the ring method of testing described below.

To apply the test, take equal parts of C and the urine. Shake thoroughly, and add aqua ammonia in excess, allowing it to run gently down the tube so as to overlay the mixture below. If the reaction be present, a deep-red band appears at the line where the ammonia meets the mixture, and when shaken, it yields a pink or rose-colored foam. To the so-called secondary reaction -i. e., a green precipitate after several hours-no further reference will be made.

The following rules must be carefully observed:

1. Use fresh urine.

2. See that the reaction is acid and the urine filtered.

3. Use a fresh test-solution.

4. Keep the sodium nitrite solution in a black bottle and in a cool place, and renew it frequently. The sulphanilic solution keeps indefinitely.

5. Hold the tube near, but not against, a white background, the source of light being behind the observer. Artificial light

should not be used.

errors.

Pseudo

reactions.

6. Accept no color but a distinct red, and regard no reaction Common as a true one in which the solution when shaken does not yield a pink foam. Pseudoreactions occur in which the band is of the proper color, but the foam is yellow or brown. The most absurd errors have arisen from a failure to observe the exact technic here outlined, and some excellent men have used sodium nitrate, weak ammonia, or even omitted the addition of ammonia altogether, and naturally failed to get any reaction at all. Others have admitted yellow and orange reactions, or have even used a 5 per cent. solution of sodium nitrite, and consequently obtained the reaction in every urine.

Value in diagnosis.

The test is not pathognomonic, as was originally maintained by Ehrlich, but is constant in all severe forms of typhoid, appearing sometimes as early as the fourth or fifth day, though more generally at the end of the first week or ten days, and persisting until the fever begins to decline. If the test be applied according to the author's method, it is absent in malaria, appendicitis, pneumonia, and the earlier stages, at least, of acute miliary tuberculosis, occurring only, in the author's experience, in some of the exanthemata, in certain cases of advanced malignant discase, and in febrile cases associated with septic absorption. Pseudoreactions are found in a considerable number of diseases.

It is consequently of great value, not alone as a positive sign, but still more as a negative one, for it is the author's firm belief that it will be found to be present at some stage in all severe cases of typhoid, and that its persistent absence in any such case quite certainly negatives the diagnosis of typhoid.

THE INSURANCE OF SUBSTANDARD LIVES.

All insurance companies would welcome any safe and equit- Desirability. able plan under which insurance might be granted along broader lines; indeed, life insurance can hardly be considered as fulfilling its whole function when it offers protection only to the families of healthy men and fails to provide for those who, being dependent upon impaired lives, have much greater need of its

benefits.

This idea has governed the action and molded the policy of several foreign companies for many years, but in the United States it has, in the past, met with little favor, though certain slight degrees of impairment have been covered by endowment policies. Quite recently one of the largest American companies has commenced to insure substandard risks of nearly all grades, and it seems probable that in the near future a similar policy may be adopted by other companies.

What is the Standard Life ?-The moment that a company attempts to insure substandard lives it becomes necessary to consider just what is meant by a standard life.

Expectation of

A standard life may be considered as one represented by an applicant who may fairly be expected to live to very old age. Nothing is more erroneous than the assumption that any man who may seem to have a fair chance of reaching his expectancy is a standard risk. The expectation of life represents merely the life not the test. mean after-lifetime of a group of men of the same age. To make good the average, there must be a certain number falling short of the mean after-life, and a certain number who exceed it. That is to say, the total ages of those who die short of their expectation must be balanced by a certain number whose total ages exceed it.

Let it be assumed, for example, that a man aged twenty-five applies for insurance; by the tables he would have an expectancy

Medical fore

casts.

Actuarial forecasts.

Statistical basis required.

of about thirty-eight years, or would reach the age of sixtythree. But if his condition indicates that he will probably live only to that age, he is manifestly not a first-class risk, for sixtythree is not old age. The question must always be, Has the individual a fair likelihood of living to a good old age, or will a thousand such live, on the average, to the age of sixty-three years? If the person looks as though he would live only to age sixty-three, he belongs to a class whose average life will be only about twenty years.

The Crucial Point.-The greatest difficulty lies, not in the mere rating of risks upon a basis of safety and profit for the insurer, but rather in making such ratings equitable and fair to all who insure.

Nothing is more uncertain, no branch of medicine more difficult, than prognosis, yet for the medical officer the whole question of insuring substandard lives is one of prognosis, and it follows that whatever system of insurance is adopted must provide for an increase in premium sufficiently great to protect the company from the element of loss due to inevitable and unavoidable

error.

The only possible way to reach a just and correct system is to make prognosis a matter of actuarial computation through the thorough and systematic collection and classification of large groups of risks of known degree and kind of impairment. If this should be found practicable, and if all companies could join hands in working out this problem, great results might be expected, and the benefit conferred would not be confined to life insurance, but would, in a large measure, accrue to the whole science of medicine. The most rational method would seem to demand an investigation of the medical history or after-lifetime of rejected risks, combined with a similar investigation of the status of accepted risks showing minor degrees of impairment. For fifty years our life-insurance companies have been rejecting from 10 per cent. to 15 per cent. of all applicants; therefore, for every ten men now insured there is at least one who has been declined and might be made a factor in such research.

Practical Difficulties.-The practical difficulties in the way are great, and by many are deemed insurmountable. Such are: (a) The fact that the class of men applying for insurance as

healthy lives and found to be unsound may be considered as representing the most favorable class of unsound lives, inasmuch as they are usually quite unaware of any serious physical defects; for as heart lesions or cases of nephritis are likely thus to be detected in their incipiency, too long an after-lifetime may be figured for such risks. On the other hand, if it becomes generally known that life-insurance companies accept impaired lives, the tendency will be to flood the offices with the applications of men who know themselves to be unsound, and are, on the average, more advanced and confirmed in their disease, and less amenable to treatment, than those upon whose after-lifetime the new mortality tables would be calculated.

(b) The elaborate and complicated system of subclassification necessary to equitable rating. It would be manifestly unfair to rate all cases of the same disease, or even the same lesion, alike, for one individual with a mitral regurgitation may live sixty years, and another die in a fortnight. And so, also, in the case of coexisting disease secondary to, or independent of, the primary lesion, subclassification of an elaborate kind would be necessary. Every degree of impairment through heredity would need to be separately considered, as well as the modifying effect of physique, age, occupation, and environment.

(c) The difficulty that must inevitably be experienced in obtaining from examining physicians a true description of the case. This source of vitiating error is absolutely unavoidable under the prevailing system. The average physician may be an excellent examiner and a man of good judgment, but to say that he can elicit or express in definite terms the finer shades of diagnostic signs that enter so greatly into prognosis involves a palpable absurdity. The expert diagnostician is frequently at his wit's end to read the future in physical signs, and it can not be expected that the general practitioner shall do more.*

(d) The question as to whether such insurance can be made at

* Dr. Glover Lyons, of London, has pointed out the fact that the medical man can deal in generalities only, and that it must always require actuarial computation to fix the probable duration of life. The physician can recognize impairment and determine the question of increasing, diminishing, or persistent hazard; he can roughly estimate the degree of impairment, but statistical methods must inevitably be called in if equitable rating is desired.