as great as that of the atmospheric lamp. They found, however, that deterioration was extremely rapid in the vacuum lamp during the first 300-400 hours of use, after which deterioration became less rapid. By the time the lamp had been 400 hours in use its intensity was reduced from about 230 units to 40 units. In the atmospheric lamp deterioration also took place, but much more slowly and more regularly, the reduction being from about 60 units to 30 units in 400 hours. These figures cannot be regarded as more than very rough guides to deterioration, as some lamps deteriorate quicker than others owing, it appears, to variations in the quality of the quartz.

Experience has shewn that for public clinics, and especially where treatment is being undertaken by inexperienced workers, a mercury vapour lamp of high intensity is neither desirable nor necessary. The output of the vacuum lamp when new is excessive, and the risk of overdosing or burning is greater than with the atmospheric type. Moreover the vacuum lamp has the disadvantage that its rapid deterioration makes dosing difficult to regulate with any degree of accuracy. The atmospheric lamp, which is less intense and deteriorates more slowly and steadily, is to be preferred. Further, this atmospheric lamp is stouter, less easily damaged, and is easier and cheaper to clean or renew. In consequence of the deterioration that all mercury vapour lamps undergo, it is important that the output of ultraviolet should be tested periodically so that the time-exposure or the distance of the patient from the lamp may be properly adjusted from time to time.

3. Methods of Estimating Output.-Several methods of estimating the ultra-violet output of lamps have been recommended. Reference was made in the first report to the acetone methylene blue gauge of Professor Leonard Hill. This gauge has been used in some centres, but the general impression is that it is not sufficiently delicate. King's carbon-tetrachloride actinometer, Levy-West pastilles, and the Hanovia Erythema dosimeter are said to be useful. These are all colorimetric tests, and while they may be helpful in determining the output of lamps from time to time, when the test is carried out by the same person, an apparatus like the photo-electric cell, which is not in any way liable to be affected by the personal equation, would appear to be more accurate. A serious disadvantage is its expense, but as it is required only at intervals, one cell might conceivably be made available for a number of centres. Other tests besides these have been devised and recommended, but it is hardly necessary to enumerate them. As the output of all forms of lamp varies considerably from time to time, it should never be taken for granted that the output remains constant. The eye is not a reliable judge of ultra-violet intensity, as the rays are invisible, and to judge of the ultra-violet output from the intensity of visible light may be quite misleading.

4. Dosage. What constitutes a suitable dose of ultra-violet?

If this question could be answered specifically most workers in ultra-violet would be grateful. The multiplicity of factors that have to be considered makes it unlikely that the dose can ever be precisely laid down. If it were practicable to say that a certain number of rays of stated wave-lengths should be allowed to act upon a certain area of normal skin for a specified period at a definite distance the dose would be brought within the region. of scientific accuracy. Professor Hill has already adopted what he calls the " I.K. unit," but except under laboratory conditions measurement of the unit is inconvenient, and it is hardly applicable for general clinical use. Once it has been determined precisely what particular wave-lengths are biologically important in treatment, and a simple and reliable method of estimating the total energy of the rays in question evolved, the fixing of a unit should be simple enough. Having determined this unit, and knowing the unit output of the lamp in use, it is merely a question of determining how long the exposure should be on normal skin. The distance of the patient from the lamp and the skin area (in relation to total area) that should be irradiated necessarily come into the calculation. If dosage could be regulated by some such method we should then be in a position to study with greater profit the varying results that are met with by different workers. Short of definite knowledge as to what irradiation had been given in each case the comparison of results is unreliable. There remains of course the responsiveness of the individual. But degrees of natural pigmentation might be arbitrarily standardised, and the few light-sensitive individuals met with would not seriously cloud the picture.

5. Erythema. In the absence of a more scientific method, testing the erythema dose of individual patients and keeping the dose of radiation just short of that necessary to produce erythema has been widely used; yet the erythema dose is not always reliable. Children up to the age of three or four years do not develop erythema readily, and some workers hold that to rely upon the erythema test as a measure of the dose in young children almost certainly means overdosing, and poor results in consequence. Eidinow has laid stress upon the value of irradiation being given always upon normal skin. The method he advocates is to expose to the rays an area considerably less than one-quarter of the body surface for a period sufficient to produce a mild erythema. At the next irradiation another area is exposed. He works on four areas, the chest, the back, the front of the legs and the back of the legs. By irradiating one of these areas every alternate day, time is allowed for the part first irradiated to become normal before the next dose is applied to it. The method not only delays the development of pigmentation but retains the skin in a state of sensitivity, and it is seldom that exposures need to be extended. There is thus a saving of time and greater accuracy in the dosage. The method has been practised to some extent in Scotland, but experience of its value has been too limited for any definite conclusions to be

formed. Some clinicians in Scotland take the view that, where the carbon arc is used, very satisfactory results are obtained by giving doses that will not produce erythema, and continuing the treatment throughout without causing any perceptible redness of the skin. In general the erythema index appears still to be the main factor on which time-exposures are based, and it is on the whole a reliable guide except in the case of young children. factor of some importance is the area of skin which is rendered erythematous, but this will be more conveniently dealt with under the next heading.

6. Bactericidal Power of the Blood.-Hill and Eidinow drew attention to the fact that when the bactericidal power of the blood was below normal it could be raised considerably by ultraviolet irradiation of the skin. When an overdose of irradiation is given the bactericidal power appears to fall off again. They regard the optimum dose as that which raises the bactericidal power to its highest level, and the erythema dose has been related to the bactericidal power by a series of careful investigations. Skin areas of different extent were tested, and it was found that for the average adult an area of approximately 26 inches square irradiated to the extent of mild erythema gave the best results. It is assumed in these investigations that the degree of improvement in the bactericidal power of the blood is a reliable index of the dose that is best for the patient. The results would appear to shew that this is the case.

1

On the assumption that the raising of the bactericidal power of the blood would probably help in destroying infection and preventing the development of infectious disease, attention has been given by some workers to the spread of infectious ailments among children who were being irradiated. The observations have shewn that irradiation has no effect in preventing infected children from developing illnesses such as "colds," chicken-pox, scarlet fever, &c. But many children apparently in robust health fall ready victims to certain infections while others less well-favoured escape or suffer a mild attack. All that is claimed is that irradiation increases the non-specific bactericidal power of the blood, and it is perhaps expecting too much to hope for the creation of specific immunity. Immunity itself is as mysterious as the action of "light," and the most that might be expected from irradiation would be a greater general resistance which prevented the development of a severe attack of illness. Whether on the whole irradiated patients generally have fewer severe cases among them than those not irradiated would seem to be an important matter for investigation. Experience along those lines has been too limited to enable any final opinion to be expressed.

7. Pigmentation.-Widely different views are held on the value of pigmentation in relation to the ultimate progress of the patient. Rollier attaches great importance to it, and considers that the cases that pigment well are the cases that do best in

tuberculosis. Observers in this country have not been able to trace any very definite relationship between pigmentation and progress, and generally it appears that light clinicians pay little attention to it. One observer in Scotland noted, in patients treated by heliotherapy, that pigmentation was often delayed until there was evidence of some improvement in the local tuberculous lesion; when improvement began pigmentation developed. But which was cause and which effect is doubtful. In many of the Scottish institutions very satisfactory results have been obtained in patients that shewed no pigmentation at any time. If our knowledge of the functions of melanin is correct, it is a little difficult to understand why pigmentation should have any relation to progress. Evidence is steadily accumulating to shew that it has probably no such relationship.

8. Constitutional Signs.-Since the erythema test cannot always be safely relied upon, and the bactericidal test is inconvenient clinically, it is well to have other guides to dosage. Initially the erythema dose should be determined and irradiation should be begun carefully with a dose below-and in the case of young children well below- that required to produce erythema. The reaction to the first treatment should be carefully noted. Normally patients feel better, or at least should not be upset. A tired feeling, headache, giddiness, tachycardia, rise of temperature, anorexia or sleeplessness are or sleeplessness are symptoms pointing to overdosing. Where any of these signs is present the patient should be allowed to recover before being again irradiated, and then a reduced exposure should be given or the distance from the lamp increased. A longer exposure to a low intensity of radiation is probably better tolerated than a shorter exposure to high intensity. In infants and young children the constitutional signs should be carefully watched. In practice the general reaction of the patient is probably the safest and best guide to dosage. Whatever method of estimating dosage is adopted, a strict watch should be kept for any signs of constitutional disturbance and the dose corrected accordingly.

9. The Carbon v. the Mercury Arc.-The carbon arc differs from the mercury arc in two main respects. It gives out more heat rays, and it is much less intense in the shorter wave-lengths of the ultra-violet. Generally the exposure required at a mercury arc is of shorter duration than at a carbon arc. As the mercury arc gives off little heat patients sometimes complain of cold, but this can be obviated by keeping the room temperature at 65° or 70° F., or by using bowl electric radiators directing the heat on the patient. The dilatation of the skin vessels by heat is said to enhance the benefits of irradiation. The carbon arc is, on the whole, more agreeable because of the heat it gives off. The fact that the presence of heat rays has an interfering effect appears to be of less importance clinically than the advantage gained by dilatation of the skin vessels and the greater comfort afforded to the patient. By compensating for the

deficiency of heat rays from the mercury arc by one or other of the methods mentioned the relative values of the two types of lamp are more nearly equalized.

In the treatment of tuberculous patients opinion appears to favour the carbon arc, but the results recorded by different workers seem to indicate that there is probably less difference between the two types of lamp than is believed. The results may be different when the patient is exposed for a longer period to radiation of low intensity than when exposed for a shorter period to radiation of high intensity. At all events investigation into the action of X-rays on cells has shewn this to be the case, and it may be that similar effects are produced in ultra-violet irradiation. If this is the case it may explain the varying results recorded in the treatment of lupus. Some light clinicians hold that the general light bath (one hour's exposure to 30 ampère lamp) does not cure lupus, whereas Sir Norman Walker obtained very satisfactory cures in a large number of cases by long exposure to lamps of low intensity (four hours' exposure to 10 ampère lamps). On the other hand, good results have been obtained in lupus cases treated by general irradiation with mercury arcs. Whatever the truth of the matter may be, there is this to remember. There are no X-rays in man's natural environment, consequently treatment by X-rays is on a somewhat different plane from treatment by ultra-violet rays. Man is capable of adjusting himself successfully to environments in which the natural intensity of ultra-violet radiation varies enormously. It takes time to obtain a satisfactory adjustment, but once attained, he reacts satisfactorily and his health is not impaired-it may be definitely improved. It would seem therefore that an important factor in artificial irradiation lies in securing proper physiological adjustment to increasing exposures. Whether the constitutional signs of over-irradiation already referred to are a sufficiently delicate test of a patient's adjusting mechanism is uncertain.

When the varying results in patients with similar lesions treated at the same lamp are contrasted with the good results obtained in other similar cases, whether treated by the carbon arc or the mercury arc, it would appear to indicate that there is something in the technique of irradiation that we do not yet fully understand.

On the whole the relative merits of the carbon and mercuryvapour arc remain undecided. The matter is of importance from an administrative as well as from a clinical point of view. The initial cost of a mercury arc where direct current is available is lower than that of the carbon arc, and the maintenance and running costs are very much less. Moreover, as the period of exposure is shorter, time is saved, and a greater number of cases can be got through with less demand on the time of the medical and nursing staff.

In the sanatoria and in one or two tuberculosis dispensaries the carbon arc is the lamp in most frequent use. But the mer