hefner unit, or hefner candle

A unit of luminous intensity, late 19th – early 20th centuries, equal to the horizontal intensity of the light from a lamp developed by the German engineer Friedrich Franz von Hefner-Alteneck (1845-1904) in 1884. Abbreviation, HK. Mainly used in Germany before 1942. One hefner unit is approximately 0.903 candela.

The lamp, known as a Hefner lamp, burned amyl acetate (C5H11C2H3O2). The wick was solid, completely filling a tube with an inner diameter of 8 millimeters. Incorporated in the lamps was an optical device for measuring precisely the height of the flame, which was to be 40 millimeters. The chief source of lack of reproducibility was humidity, the lamp being about 10% brighter in dry air than at high humidities, which was corrected for by standard tables. In Germany the hefner unit superseded the vereinkerze for scientific work, and was replaced by the candela.

Friedrich Franz von Hefner-Alteneck.
Vorschlag zur Beschaffung einer konstanten Lichteinheit (“Recommendation for provision of a constant light standard”).
Elektrotechnische Zeitschrift, volume 5, pages 20-24 (1884).


Photographs of two 1904 Hefner Lamps “approved by the Physikalish-Technichen Reichanstalt as the German Photometric Standard Source of Light, burning amyl-acetate”, can be seen at http://physics.kenyon.edu/EarlyApparatus/Optics/Lamp/Lamp.html, courtesy of Professor Greenslade.



Eitner's lamp, as well as all those which are based on the combustion of mineral oils, is affected by the same cause of error. These liquids are not of well-determined chemical composition, but are mixtures of different substances having different boiling-points and variable compositions; they cannot be obtained in conditions which are always identical. They have further the disadvantage of not burning uniformly, the combustion at first being at the expense of the most volatile materials; there remains finally a product volatilizable with difficulty, which requires other conditions of combustion to furnish the same flame.

These considerations and a great number of tests of different liquids induced von Hefner-Alteneck to take as a combustible acetate of amyl. This liquid is fluid and possesses a very intense, agreeable odor. It may be easily obtained pure by distilling crystallizable acetic acid, or an acetate, with sulphuric acid and amylic alcohol. It is manufactured in great quantities for perfumery; its boiling-point is very constant at 138°C., and its price is not high.

The Hefner Lamp.

90. The von Hefner-Alteneck lamp is a simple spirit lamp (Figs. 43 and 43 bis); the inventor retained the wick because the lamp is manipulated more easily, and because, further, the wick does not char in burning acetate of amyl; its object is, in fact, simply to suck up the liquid which is disengaged as vapor when the temperature reaches 138°C.

The wick-holder is a German silver tube, 8 mm. in interior diameter, 0.15 mm. thick, and 25 mm. high. The normal intensity of the lamp is determined by the height of the flame; this height is normally 40 mm., or five times the diameter, measured from the top of the wick-holder; it is regulated by raising the wick more or less in the latter. The normal height is kept by means of a sight fitted to the lamp.

The flame should burn freely in the air, without a glass chimney; however, a straight glass tube is sometimes used, 88 mm. in height, and 55 mm. in diameter; under these conditions the luminous intensity of the 40 mm. flame diminishes 2 per cent.

The wick should be made with great care; it should exactly fill the German silver tube without being crowded. One may make it himself by placing parallel to one another ordinary cotton threads, until the required diameter is reached. It is not advantageous to employ, as has been proposed, a wick whose end is made of threads of amiantus [a fine kind of asbestos]; the complication which results is not compensated by the slightly greater uniformity of the light thus obtained; and, further, it is not proved that the intensity of the latter is unaltered; moreover, the amiantus does not at all dispense with cutting the wick from time to time.

The normal height of flame of 40 mm. was adopted because the lamp then gives a light equal to that of an English candle whose flame is 43 mm. in height. However, Bunte† found that the von Hefner-Alteneck unit corresponds to the English candle with a flame 45 mm. in height, and Liebenthal concluded, from more than 200 comparisons, that the flame of the acetate of amyl lamp, 37 mm. in height, has the same luminous intensity as that of the English candle 43.2 mm. in height. These differences are not exaggerated if we take account of the want of uniformity of these two light standards, and the difficulties of measuring with exactness the height of the flame.

Some comparisons, made with the greatest care with perfected apparatus, have been effected by the commission on photometric standards of the German Society of Gas Engineers, and by Lummer and Brodhun, of the Physico-Technical Institute of Berlin.

Below are the results obtained, which are the mean of a great number of measurements made with different lamps and candles.

1 German candle equals 1.223 Hefner units (German Commission).
                                            1.162    "        "    (Lummer).

1 English candle equals 1.129  Hefner units (German Commission).

The difference of 6 per cent between the results of the commission and those of Lummer and Brodhun shows well the difficulties of obtaining photometric standards with free combustion.

The intensity of the Hefner standard is too small, especially now that the tendency is more and more toward employing intense radiants; the color of the flame is somewhat red, on account of its relatively low temperature. It is the richest in red rays of all light standards. On this account, also, we cannot use it with advantage in the photometry of incandescent and arc lamps.

91. The sole advantage of the acetate of amyl lamp is to be found in its great constancy; on this point all those who have used it agree. Thus, Liebenthal determined from a great number of measurements that the mean error of one observation varies between 0.0 and 1.5 per cent. In 225 observations by Dibdin with the Hefner lamp, the result differed from the mean by a quantity less than 1 per cent in 206 (90 per cent of the measurements).

* Lum. Él., Vol. X. p. 601; Electr. Zeitsch., Vol. IV., 1883; Vol. III. p. 20, 1884.

Journal für Gasbeleuchtung, 1885, p. 796.

A. Palaz.
A Treatise on Industrial Photometry with Special Attention to Electric Lighting.
Authorized translation from the French by George W. Patterson, Jr., and Merib Rowley Patterson.
New York: Van Nostrand Co, 1894.
Pages 135-139. The first French edition, Traité de Photométrie Industrielle Spécialement Appliquée à L’Éclairage Électrique, was published in 1892.

home| units index| search| to contact Sizes drawing of envelope| acknowledgements| help|


terms of use