While electric lamps both of the cap and hand type are being introduced into many mines requiring the use of safety lamps, the oil-burning safety lamp is still used in the great majority of cases, and even where the former are used the latter must continue to be employed when testing for gas. It is not the intention to discuss here the relative merits of the two types, but simply to give a comparison of the illuminating effect or candlepower of several kinds of ordinary safety lamps used for general work and for testing. Few data have been published in recent years on this subject and because of this it is hoped that the following information may be of some interest and value.

The lamps tested were in all cases of standard size and in good condition, all practically new, and were selected from a collection of about 100 with a view to obtaining results on as many different common types as possible. With the exception of the gasoline-burning lamps, five different oils were used in each lamp to determine whether particular oils were suited to particular lamps and whether there was any great difference in the illuminating power of the different oils, or whether some oils would maintain their illuminating power better than others throughout a working period of 8 hr. No attempt was made to determine the amount or cost of oil burned by each lamp, but the condition of the lamp, wick, glass, and gauzes was noted after each test.

Description of Apparatus Used

The candlepower was determined with a Weber photometer, Fig. 1, which consists of a tube A containing a movable milk-glass screen adjustable by the milled screw B. The distance of this screen from the standard light at C is read on the scale attached to A. The standard light consists of a benzine, lamp so arranged that the height of the flame may be adjusted exactly by observing it against a scale in the lamp box through an opening in the front. The lamp to be tested is placed at D so as to illuminate a milk-glass screen in the holder at E. At the intersection of the axes of tube A and sight-tube F are placed two 45° prisms with their surfaces in contact so that the light from D may pass through without deflection while that from C is reflected at the contact in a direction parallel to that from D. The faces of the prisms are blackened around an elliptical field and on this field is an annular silvered ring which reflects the light from C. When viewed through the telescopic eyepiece, G, focused on the prisms, on outer annular ring and inner elliptical surface are seen illuminated by light D and an inner ring illuminated by that from C. With D in a fixed position, the milk-glass screen in A is adjusted until an evenly illuminated field is observed through the eyepiece.

Then the screens in A and E are illuminated to the same intensity and the intensities of the standard and unknown light are to each other as the squares of their distances from the respective screens, or expressed as a formula

i = R²/r² C

in which i = candlepower or intensity of light being tested,
R = distance of D (light being tested) from screen E,
r = distance of C (standard light) from screen in tube A,
C = constant involving the candlepower of the standard lamp, relative opacity of the two screens and characteristics of the apparatus.
Determination of the constant C was accomplished by placing a standard candle at D. The value of C is then given by writing the formula

C = r²/R²

as the intensity i is then unity. About 50 readings were taken using standard spermaceti candles. Several candles were employed and their illuminating power checked by weighing before and after use and determining the consumption of spermaceti per hour. A standard candle is supposed to burn 120 grains per hour. The candles used consumed from 123 to 124 grains per hour equivalent to a candlepower of 1.03. The average value obtained of the constant. C was 0.36. The candle was placed at the same distance as the lamps tested and also at greater and less distances.

The specific gravity of the oils was determined with a Westphal balance, and the Baume degrees calculated.

Description of the Miner’s Lamps

The lamps tested were selected to give a wide range in variety and were of standard make and size (Figs. 2, 3 and 4). Before beginning each test with the several oils they were all thoroughly cleaned, the oil vessels rinsed out with gasoline and dried and fresh wicks put in. The shape and size of wicks, number and dimensions of gauzes and oils used are given in Table 1.

Lamp No. 1 is an ordinary Davy with aluminum frame.

Lamp No. 2 is a Davy lamp with bonnet and glass shield surrounding the lower portion of the gauze but which could be lifted when testing. It is also known as the Hughes-Evan Thomas lamp. Normally it takes its air through openings in the base ring covered on the inside by the lower end of the gauze.

Lamp No. 3 is of the same type but with aluminum oil vessel, frame, and bonnet. Also it has a Hailwood porcelain flat wick burner.

Lamp No. 4 is an ordinary improved Clanny unbonneted with brass oil vessel and frame.

Lamp No. 5 is constructed on the principle of a Clanny, unbonneted, but the gauze consists of perforated sheet copper instead of woven wire.

It, is also equipped with a Hailwood porcelain flat burner and brass oil vessel and frame.

Lamp No. 6 is a bonneted Clanny very similar to the lamp known as the Evan Thomas No. 7. The air is admitted through openings in the

lower part of the bonnet, and a deflecting ring is placed between these openings and the gauze, requiring the air to pass upward and over this deflector. The oil vessel and frame is brass and bonnet is of sheet iron.

Lamp No. 7 is a standard Hailwood working lamp constructed on the Marsaut principle with two gauzes. Oil vessel and frame of brass, bonnet of sheet iron, and porcelain flat-wick burner.

Lamp No. 8 is a bonneted Clanny with smoke gauze.

Lamp No. 9 is a Marsaut type with Stokes shut-off ring at base of bonnet by which air is admitted at that point or, when closed, forced to enter the lower set of holes near the top of bonnet and pass down behind a deflecting ring extending from just above the inlet openings to about the middle of the bonnet. The lamp is fitted with Hailwood flat-wick porcelain burner.

Lamp No. 10 is a Beard Deputy with Beard-Mackie sight indicator. The circulation of air is on the Eloin principle, the air entering through gauze-covered openings around the base below the flame.

Lamp No. 11 is a testing lamp of English manufacture. By means of a sliding ring around the base the air may enter gauze-covered openings below the flame, or, when these are closed, it enters through openings in the lower part of the bonnet passing through the lower part of the gauze as in the ordinary Clanny lamp.

Lamp No. 12 is an ordinary Ashworth-Hepplewhite-Gray.

Lamp No. 13 is an improved Ashworth-Hepplewhite-Gray with Hailwood porcelain flat-wick burner and six hollow inlet tubes from fop to bottom, each with an opening at the bottom, instead of four inlet tubes as in the ordinary form No. 12.

Lamp No. 14 is an improved Gray lamp with brass flat-wick burner. A single hollow standard of large size connects the gauze-covered air inlet in the base with the interior of the bonnet. This standard has an opening near the bottom with sliding sleeve to cover it. The gauze is of much larger diameter than those of Nos. 12 and 13.

Lamps No. 15 and 17 are respectively round and flat-wick Koehler’s, burning gasoline.

Lamp No. 18 is a flat-wick Cremer, burning gasoline, differing from the Koehler and Wolf in the construction of the bonnet. The, air enters through slots in the circumference of the base of the bonnet and not through perforations along the sides. It has a corrugated expansion ring between the base of bonnet and flange of the gauzes similar to the Koehler.

Lamps No. 19 and, 20 are respectively standard round- and flat-wick Wolf’s.

The five gasoline lamps are all fitted with Pyro igniters.

The illuminating power of some of the types tested as quoted by other authorities is given in Table 2. A wide variation in some cases will be noted., It may be proper to mention at this point that some of the results obtained were checked by testing the same lamps with a Bunsen type photometer, the results agreeing closely. The greatest difference will be noted in Nos. 10 and 12. No. 10 was an aluminum Beard-Deputy with Beard-Mackie sight indicator. To check results on this, No. 10a was tested with one oil. No. 10a lamp was identical except that it was made of brass, the indicator was omitted, and it did not have a bonnet. The results on No. 12, A-H-G lamp, were checked with a duplicate lamp, the results agreeing within 5 per cent. The good lighting power of No. 13 over No. 12 is undoubtedly due to the better air supply of No. 13. As noted in Table 1, this lamp has six air tubes each with an air hole at the base, whereas No. 12 has four tubes with air inlets near the base in two only. A considerable difference will also be noticed between the observed candlepower of the Wolf type lamps and the generally accepted candlepower with a value of 1.

Conditions of Test

After being cleaned, filled, and lighted, the flame of each lamp was adjusted to a normal height suitable for working conditions, as near as it was possible to judge. This was ordinarily done by turning up the flame until it would begin to smoke and then slowly lowering it until it stopped smoking. The lamps were then allowed to stand 20 to 30 min. or until all parts were thoroughly warmed up, before the first readings were taken. The flames were again adjusted just before the readings were made. They were also maintained at approximately a working height throughout the time of 8 hr. and carefully adjusted at each observation.

Illuminants

The oils used were Eureka Safety Lamp oil, Mineral Sea oil, pure sperm oil, pure lard oil, a mixture of lard oil and kerosene, and gasoline. The first two are mixed oils sold on the market under the trade names given. Eureka Safety Lamp oil is apparently a mixture of lard or cotton seed, possibly both, and a petroleum oil. Mineral Seal is a mixture of fish or sperm oil and petroleum oil with possibly other ingredients. The mixture of lard oil and kerosene was prepared for the purpose of this test and was composed of 50 per cent, lard oil and 50 per cent, kerosene by volume. Pure lard oil was tried in all the lamps, but even after they were warmed up such constant attention was required to keep them burning properly that any results obtained would have been of no practical value.

The specific gravities of the oils and the corresponding gravity on the Baume scale are given in Table 3.

Conclusions

The average candlepower for the different oils indicates clearly the advantage of a mixed oil or the addition of petroleum oil. The average illuminating powers of lamps Nos. 1 to 14 for the different oils, excepting gasoline, are given in Table 4.

There is apparently no definite variation of candlepower with the specific gravity of the oil, for although sperm oil is the heaviest and has the lowest average illuminating power, Mineral Seal, the lightest, is intermediate in illuminating power between sperm and the other two heavier oils.

In addition to better lighting, the mixed oils burned more freely and required less attention to the wick and flame, but showed a slightly greater inclination to smoke than the pure sperm oil. The better illuminating power of the mixed oils is especially noticeable in Lamps Nos. 4, 6, 10, and 11. The effect of a bonnet in decreasing illuminating power is indicated to some extent in Lamps Nos. 10 and 10a, identical except for a bonnet. With the mixed lard and kerosene a difference of 0.04 candlepower is noted.

The effect of admitting the air below the flame or even at the base of the bonnet, instead of higher up, is shown conclusively in lamps Nos. 9 and 11, constructed so that air could be admitted in different ways. In No. 9, the results given in Table 1 were obtained with air entering at base of bonnet; the same lamp with air entering near the top of the bonnet and passing downward between bonnet and deflecting ring gave an average candlepower of 0.24 with Eureka oil, 0.07 less than the value given in Table 1. With No. 11 the results are even more pronounced. With the inlet at the bottom closed, and air entering at base of bonnet and gauze, the average candlepower with Eureka oil was 0.23, while with air entering below the flame it was 0.36, a difference of. 0.13 or 36 per cent.

In testing the lamps with flat wicks, the candlepower was in all cases taken in a line normal to the width of the flame. The decrease in lighting edgewise to the flame was much less than anticipated. For the oil lamps the candlepower edgewise varied from 5 to 20 per cent, less with an average of 10.5 per cent. For the gasoline lamps the illumination edgewise was 4 to 7 per cent. less.

With the exception of gasoline, the decrease of candlepower during the 8 hr. seems to vary inversely as the illuminating power of the oil. Sperm oil gave an average candlepower for 14 lamps when freshly lit of 0.301 and after 8 hr. of 0.304, again of 1 per cent.; Eureka oil gave at the beginning an average of 0.323 cp. and after 8 hr. 0.308 cp., a decrease of 4.3 per cent.; Mineral Seal at beginning an average of 0.340 cp. and after 8. hr. 0,314 cp., a decrease of 7 per cent.; a mixture of 50 per cent, lard and 50 per cent, kerosene at the beginning gave an average of 0.403 cp. and after 8 hr. 0.296 cp., a decrease of 26.5 per cent. To determine whether the decrease in candlepower was due to the lighter part of the oil burning off first, leaving a heavier residue, the specific gravity of the oil remaining in the lamps after burning the mixture of kerosene and lard oil was determined, This was found to be 0.865 or 0.005 greater than the original oil. As this is only 0.58 per cent, gain, it is within a possible experimental error, and it would be hardly safe to assume from it that there was a separation of the oils. It would be more reasonable to conclude that the decrease in illumination was due to the more rapid deposit of soot on the gauze and a film of unburned carbon on the glass when burning oils containing petroleum products, than when burning pure sperm oil.

With the gasoline lamps, the illuminating power decreased in 8 hr. from an average of 0.65 to 0.62 cp. or 4.6 per cent.