years. In the United States the 230-volt. 3-wire system is the rule, but in Europe, and especially in Great Britain, the 450-volt, 3-wire system is equally the rule. One system employs 115-volt lamps on each side of the neutral, and the other requires 225volt lamps. It is obvious that copper economies are with the 450-volt system, but on the other hand, the lamp efficiencies are with the 230-volt system. In Europe, moreover, the standard is 8-cp instead of 16-cp, and it is the fact that here again the lamp efficiencies are with the 115-volt lamp of the higher candlepower. It would seem, therefore, that there is no immediate prospect of a change of American distributing circuits from 230 to 450 volts on the 3-wire distributing networks, but that on the contrary, the new lamps may find a wider market here. while confirming the practice at 115 volts.

Discussing recently before the Iowa Electrical Association the selection and design of a distributing system from the standpoint of cost and operation, Professor A. H. Ford, of the State University of Iowa, expressed the opinion that when the load is concentrated within approximately half a mile of the power station and there is the possibility of a considerable variablespeed motor load being developed, it is advisable to install a direct-current 3-wire system with a potential of 115 volts on each side. By increasing the potential to 230 volts, a great saving in conductors is accomplished, but this is offset by lower lamp efficiency and increased trouble with cutouts, switches, and so forth, on customers' premises. When the load is scattered, a polyphase system should be installed-2-phase, 4-wire for plants under 1500-kw capacity and 3-phase, 4-wire for plants larger than this. This transformer potential is preferably 2200 volts primary, with a 3-wire secondary having 110 volts on each side. The secondaries of all lighting transformers on each phase and feeder should be interconnected where it can be done without extending the circuit more than 200 feet. Separate transformers should be used for the motor load, though each motor need not under every condition be supplied with a separate set of transformers.

It may be noted that the question of satisfactory fuses for heavy power circuits of from 2000 to 5000 volts for use on distribution pole lines for light and power has been found serious.

and becomes more so as the number increases of customers who take large blocks of power for factory purposes from centralstation companies. While these customers if supplied by smaller stations would be important enough to justify the use of separate feeder circuits with station circuit-breakers, it becomes necessary in a large distributing system to supply many of them by means of branches from main power-distributing feeders. The Detroit Edison Company, with its large new plant at Delray, described in the Denver report on progress in 1905, has encountered and dealt with this problem. The voltage of transmission from Delray to the substations is 4600 volts, 3-phase. Large power consumers are not supplied through the substations at the usual standard direct and alternating-current voltages, but are fed from the 4600-volt lines. This brought up at once the necessity for fuses for customers' branch circuits supplied from these transmission lines. The fuse now used in Detroit consists of a 25-ampere (or less) 5000-volt enclosed fuse, held in a case or carrier which is supported on insulators on the pole and held in a vertical position. The case is a piece of fibre conduit, two inches inside diameter. Into this are expanded two brass rings, to each of which four phosphor-bronze clips are fastened. The enclosed fuse is inserted from below by means of a hard-rubber handle to which it is attached. A spun-brass cap is put over the top of the case. A number of these 25-ampere fuses have been shortcircuited on the company's 4600-volt station 'bus, and all have opened the circuit in good shape. Of the fuses that have blown in the service, a few have failed to open the circuit and have burned up. This has always occurred during lightning-storms, and the company's engineers have attributed the failure of the fuses to abnormally high potential on the lines during the lightning discharge. No difficulty has been experienced with moisture getting into the fuse tubes.

ILLUMINATING ENGINEERING

A branch of the central-station art that demands yearly more attention is illuminating engineering. If it be true that 20 per cent of the electrical energy hitherto employed in lighting has been wasted by poor methods of utilization, the study of improved principles and appliances at once justifies itself on strictly com

mercial grounds, to say nothing of all the other considerations, some technical, some economic, some physiological. The relation, for example, between illuminating engineering and free lamp renewals may not be obvious, but is a very real and definite one. Mention was made not long ago of one customer of a central-station company that did not give free lamp service. He burned all his lamps dim, one by one, moved them around from socket to socket, and finally had one good lamp left in the house, which he carried from room to room as he needed light. The desirability of free lamps in the cause of better illuminating engineering might be inferred from cases such as this.

During the past year the United States Government itself has, through its proper representatives, taken up the question of standardizing specifications for incandescent lamps; and at least one important conference has been had on the subject-that at Washington in February. Associated with this question, as a matter of general welfare, is that of keeping down in the future the number of sizes and kinds of lamps. It is easy to understand the anxiety that was felt in many quarters last winter as to an adequate supply of lamps, but it is an impressive fact as to the effect of the existing heterogeneity of demand and practice that the largest producer has to maintain the enormous stock of 78,000,000 lamps in order to fill orders for all the styles and voltages that it has to meet.

Incidental in a practical way to the question of lamp renewals, and so forth, is the simple plan noted by Dr. Clayton H. Sharp, of the Electrical Testing Laboratories of New York, of assorting lamps as to age. Heretofore it has been customary to do such assorting photometrically, which method is tedious, necessitates some photometric skill, and is expensive in application. The experiments of Dr. Sharp indicate that in assorting incandescent lamps for age sufficiently accurate results from the practical standpoint may be obtained by simply fixing the degree of blackening of the globes of lamps that have seen some service, by comparison with a set of lamp globes of graded degrees of blackening. When the blackening is observed to be identical with that of one of the lamps of the scale, it is assumed that the life of the lamp being tested is equal to the life that has been accurately determined to correspond to the degree of darkening

of the reference lamp. It is evident how greatly this method. simplifies the assortment of lamps, and the investigations cited by Dr. Sharp establish the commercial accuracy of the method for the classes of lamps experimented with. Even should a single series not apply to all lamps, the value of the method would be affected only in so far as it involved different reference series for different kinds of lamps, and even this might not be necessary for sufficiently approximate results. In fact, a very considerable range of inaccuracy would not put the method at a disadvantage with respect to the photometer method with its great possibilities of error when applied outside of a laboratory. The investigation by Dr. Sharp also brought out the interesting fact that 50 per cent of the decrease in candle-power of a lamp is produced by bulb blackening and 50 per cent by changes in the filament.

Strictly speaking, illuminating engineering, while comprehensive of such topics, has found abundant scope for all its energies thus far in the discussion of principles, appliances, fixtures, and so forth, and the fact that the new Illuminating Engineering Society should have in 1907, at the close of its first year, a membership of over 800 tells its own story as to the need for this technical organization, before which Dr. C. P. Steinmetz in December, discussing light and illumination, outlined a large field for work and usefulness. The illuminating engineer, he said, is concerned more particularly with securing illumination rather than light. The problem which confronts him is one dealing with effects rather than causes. In each specific problem it is ordinarily desirable to have both general and concentrated illumination. Thus, in an office one wishes to have the whole room illuminated at a low density and the desk brilliantly lighted. Illumination may be either direct or indirect. Although, with reference to the light rays obtained, the direct is much the more efficient, yet for almost all purposes the indirect gives the better illuminating effect. It is particularly desirable to have uniformity of illumination rather than concentration of light for the general lighting. For the concentrated lighting it is best to use a single lamp with reflectors which cause the rays to pass in the desired direction. In any event, the lamps should be placed entirely out of the field of vision. The prime requisite is to have the light fall upon the object that one wishes to see without

too pronounced shadows and without an excessive amount of light from other sources reaching the retina. The amount of light that reaches the retina depends upon the opening of the pupil, and the contraction of the pupil depends largely upon the heat energy contained in the rays that strike the eye. The eye is so constructed that by variation in the opening of the pupil the amount of light that reaches the retina may not be largely varied even when the density of illumination changes in the ratio of I to 1000. It is the task of the illuminating engineer to so place the lamps that the heat energy of the rays that reach the eye will allow the pupil to expand to such an extent that the light rays that come from the body that is to be viewed will be of just the right density when they reach the retina. Thus all lamps should be removed from the line of vision and all glare should be carefully avoided. As Dr. Steinmetz pointed out, light is not in reality a quantity to be treated by ordinary engineering methods, but, making itself manifest as an excitation of the retina, also relates rather to physiology than to physics.

Reference must be made to the specifications for incandescent carbon-filament lamps issued this year by the British Engineering Standards Committee. It is decided that the British light standard shall be the 10-cp Vernon-Harcourt pentane lamp at the British National Physical Laboratory, from which carbon filament secondary standards shall be derived. The useful life of a lamp is completed when the mean horizontal candle-power has dropped 20 per cent. Four classes of standard lamps are recognized, having lives of 400 and 800 hours, respectively, and made for 110 and 220 volts; these are distinguished by reference letters, which also indicate the standard consumptions. Thus, for 16-cp lamps, A represents a 400-hour, 50-watt lamp for 110 volts; B a 400-hour, 60-watt lamp for 220 volts; C an 800-hour, 56-watt lamp for 110 volts, and D an 800-hour, 66-watt lamp for 220 volts. The reference letter (in a circle), voltage and mean horizontal candle-power, are to be marked on each lamp bulb, with the name or trade-mark of the manufacturer. The candlepower is to be determined at standard pressure, while the lamp is revolving at 200 revolutions per minute about a vertical axis. The average limits between which the results must fall are 8 per cent for the candle-power and 5 per cent for the total watts,