H 



NA TURE 



[July 7, 1923 



scale. Thus a note of any desired pitch can easily be 

 obtained, but the intensity varies on account of the 

 varying sensibility of the ear and the apparatus. This 

 possibility of variation of pitch makes a number of new 

 methods of wireless signalling feasible. One of the 

 easiest resembles a very early kind of moving needle 

 telegraph apparatus called Bright's bells in which the 

 needle moved to one side and struck a bell in order to 

 indicate a dot and moved to the other side and struck a 

 bell of different tone to indicate a dash. This method 

 was faster than the dot and dash sounder and appar- 

 ently easier to learn. In its proposed wireless form the 

 transmitting station would emit equal wave trains to 

 represent dots and dashes, say of 200,200 frequency to 

 represent the dots and 200,500 frequency to represent 

 the dashes. Each Morse sign is then heard as a little 

 melody at a receiving station using a local oscillator of 

 200,000 frequency. Besides the advantage mentioned 

 above there is a likelihood that these signals would be 

 less distorted by atmospheric discharges than are longs 

 and shorts of constant pitch. 



Still another simple method consists in utilising 

 three very close high-frequency oscillations at the trans- 

 mitting station, say 200,200, 200,100 and 200,050, and 

 making a new code for the alphabet out of permutations 

 of these. The local oscillator would have a frequency 

 of 200,000, and therefore the sounds heard in the tele- 

 phone would be short tunes. The method would be 

 faster than Morse, but might demand that the operators 

 should have musical ears. Still another method can be 

 imagined in which chords of three notes instead of 

 arpeggios are used for the letters of the alphabet, but 

 this might require an even more musical ear. 



But there is one kind of chord which every one can 

 recognise without special training, which even the 

 horse can discriminate in the sounds of " whoa " and 

 " gee." The vowel sounds are in fact chords. Lately 

 Sir Richard Paget has given (Vowel Resonances, Inter- 

 national Phonetic Association) a list of the chief tones 

 occurring in the English vowels. For example, the 

 vowel sound in the word " calm " contains the tones of 

 frequency 1 360 and 8 1 o per second . Suppose, therefore, 

 a transmitting station is arranged to emit simul- 

 taneously electric waves of frequencies 201,360 and 

 200,810, and suppose these waves when received at a 

 great distance are combined with local oscillations of 

 frequency 200,000 per second. Then the tones 1360 

 and 810 are perceived simultaneously as a chord in the 

 operators' telephones. But this chord by itself is 

 scarcely if at all recognisable as a vowel. Recognition 

 is ensured by superposing a larynx note by aid of a 

 buzzing contact included in the receiving circuit. Then 

 whenever a train of two waves leaves the sending 

 station the vowel is pronounced by the receiving 

 apparatus. This is easily illustrated to an audience by 



the aid of a loud-speaking telephone. Lecture appar- 

 atus for producing and detecting the two vowel sounds 

 represented by 0, a, is shown in Fig. 9. The change of 



Hi-— |f 



BUZZING 

 CONTACT 



RECEIVER 



Fig. 9. — Heterodyne vowel apparatus. 



radio frequency necessary for passing from one vowel to 

 another is provided by the tappings on the inductance 

 coils. In this apparatus the transmission occurs across 

 a short distance ; in practical telegraphy the trans- 

 mitter would be more powerful and would be provided 

 with an aerial and the receiving apparatus would also 

 have an aerial. 



The apparatus, which was built and made to work 

 by Messrs. C. F. A. Wagstafife and E. S. Smith, two 

 former Finsbury Technical College students, was con- 

 structed to produce six vowels, namely, those heard in 

 the words eat, all, hate, shoe, calm, and earth. These 

 six vowels taken in pairs yield thirty-six symbols 

 which, together with the five vowels a, e, i, 0, u repre- 

 senting themselves, amount altogether to forty-one 

 symbols. An alphabet formed in this manner is much 

 briefer than the Morse code ; that is to say, there are 

 fewer efforts of the sending key in making the same 

 message. For example, in the word London there 

 are seventeen efforts when Morse is used but only eight 

 when the vowel code is employed. Besides the gain 

 in speed there is a possibility of reception through 

 atmospheric disturbances being more easily accom- 

 plished with the vowel code than with the customary 

 dots and dashes of constant pitch, but this can only be 

 tested by actual trials. 



Ur of the Chaldees. 



By C. Leonard Woolley 

 TN 1919 Dr. II. R. Hall, on behalf of the British 

 -L Museum, spent three months excavating at Ur. 

 Last summer the British Museum and the University 

 Museum of Philadelphia decided to send out a joint 

 expedition which should continue for a term of years 

 the work begun by Dr. Hall, and clear as much of the 



site as seemed likely to . repay the necessarily heavy 

 cost of a scientific mission. The first season's work 

 of the joint expedition is now over, and the results 

 amply justify the confidence of those who promoted it, 

 and give every promise of even greater success in the 

 future. 



NO. 



2801, VOL. I 12] 



