February 2, 1899] 



NA TURE 



321 



was such that the development of the under-exposed 

 plates was exceedingly laborious, and the results were 

 most unsatisfactory. He turned his attention, therefore, 

 to increasing the actinic effect by the use of a burner in 

 which the flame, at first of ordinary illuminating gas, was 

 surrounded by pure oxygen. With this form of burner, 

 a diagram of which is given in Fig. 2, photographs were 

 readily obtained upon a moving plate, in which the 

 salient features of the images described by Koenig were 

 clearly brought out. The gas was subsequently enriched 

 by passing it through a receiver of petroleum ether, and 

 in this way the brilliancy of the flame was further greatly 

 increased. In Merritt's experiments the moving plate 

 was shot horizontally through the field of the camera at 

 a speed sufficient to separate properly the various flame- 

 images. The speed of the plate-holder, which was 

 arranged to slide between guides, was about two metres 

 per second. The entire time-period covered by the 

 chrono-photographs thus produced was only a few 

 hundredths of a second. 



Chrono-photographs of the manometric flame have 

 since been made by Hallock and Muckey {The Looker 

 On, 1896, pp. 1, 177 and 375, 1896), who used such flames, 

 excited by resonators, in the analysis of the voices of 

 various opera singers ; by the writer in collaboration 

 with Prof. 'Mertln {P/tysical Review, yo\. vii. p. 93, 1897), 



and by Miss J. A. Holmes {T/iesis~m manuscript- 

 Library of Cornell University, 1898). 



Acetylene gas, which has come into common use since 

 the experiments just described were made, affords a light 

 of much greater actinic power than any which was 

 formerly available. The flame of burning magnesium 

 alone surpasses that of acetylene in brilliancy. The 

 carbon bands in the electric arc, to be sure, give that 

 source of light, likewise higher actinic value than the 

 acetylene fiame ; but the arc light cannot be used 

 manometrically, nor, indeed, is it probable that the 

 magnesium flame could be thus employed. 



When we surround the acetylene flame with pure 

 oxygen in a burner, like that described by Prof. Merritt, 

 its actinic power is still further increased. 



In 1897 the writer spent many pleasant hours of the 

 summer vacation with Prof Merritt in the fascinating 

 work of photographing the manometric flame. The 

 e<perinients of 1893 were repeated with acetylene in 

 p'ace of ordinary enriched burning gas, and with films 

 of considerable length instead of the glass plates. The 

 manometric burner was the same in all essential features 

 ■IS that described by Merritt in the article which I 

 liave just cited. It was supplied with a mixture of 

 equal volumes of acetylene gas, generated by the action 

 of water upon calcium carbide in the usual manner, 

 and of hydrogen. The chrono-photographs were taken 

 upon films 120 cm. in length, which for convenient 

 NO. 1527, VOL. 59] 



handling were mounted in an especially constructed 

 camera. This camera consisted of the usual lens and 

 bellows, and of a rectangular box of wood containing 

 a drum Ji (Fig. 3), upon the periphery of which the 

 film was mounted. The drum could be driven at a 

 convenient speed, either by means of a belt attached to 

 an electric motor, as shown in the diagram, or, as was 

 sometimes found to be more convenient, by hand. The 

 box which contained the drum was light-tight, excepting 

 that at a position suitable to allow the passage of the 

 rays from the lens there was a vertical slit closed by 

 a shutter. This shutter could be opened electrically by 

 an observer stationed at the manometric flame, after 

 which it remained open for precisely one revolution of 

 the drum. When this revolution was completed, the 

 shutter closed automatically. 



The revolving drum, which carried the sensitised film 

 upon which the photographs of the flame were taken, 

 was given a speed in most of our experiments of about 

 one revolution per second. This was found to be quite 

 sufficient for the proper separation of the flame-images, 

 and it permitted us to record upon a single film any 

 word or phrase the utterance of which did not require 

 more than a second of time. In certain cases, where 

 we desired to include in the chrono-photograph poly- 

 syllabic words or phrases, the speed was somewhat 

 reduced ; in other cases, for the purpose of a further 

 separation of the flame-images, the drum was driven. 

 at a much higher velocity. 



^ 



Fig. 3. — Camera for photographing the 



the revolving dn 



epreseiued by the dotted circle D.] 



In the manner just described a large number of striking 

 photographs were obtained, the beauty and sharpness of 

 detail of which no adequate idea can be given in the 

 printed reproduction. It was found that the repetition 

 of the same combination of articulate sounds, uttered at 

 the same pitch and by the same speaker, always gave 

 very closely indeed the same series of flame-images. 

 Nevertheless the reading and interpretation of these- 

 photographic representations of the manometric flame is 

 by no means a simple matter. When we attempt to read 

 such a record, as one would read the trace of the syphon 

 recorder in a telegraphic message, or as one would read 

 shorthand, we find that it is only the vowels which 

 produce any marked agitation of the flame. All those 

 accompanying mouth-sounds which introduce and close 

 each syllable in articulate speech, and by which, in great 

 measure, we are able to distinguish the different words,, 

 produce a very feeble and often an unrecognisable effect 

 upon the flame. The records are indeed the very 

 opposite of shorthand writing, not only in that instead of 

 a single character to a syllable, we have sometimes as 

 many as a hundred oscillations of the flame, but likewise 

 in the fact that while shorthand is made up of words 

 with the vowels left out, these manometric photographs 

 represent speech with the consonants suppressed. It is 

 obvious that to read a record of the latter sort, even after 

 the eye had been trained to recognise the flame- 

 groupings characteristic of all the vowel sounds, is more 

 difficult than it is to pick out words in which the con- 

 sonants are indicated and the vowels omitted. 



There is in the interpretation of the flame photographs- 



