October 14, 1892,] 



SCIENCE. 



2 17 



fiir das menschliche Auge in ewige Nacht gehullten Strahlen- 

 bereiohe entquellen, gar kein Ende nehmen. Mit jedei- folgenden 

 Region, die man zur Aufnahme einstellt, meinfc man das End- 

 gebiet der kleinsten Liclitwellen zu erreiohen. Aber es ist fast, 

 als flohe die kleinste Welle, die iiberhaupt noch photographisch 

 zu fesseln ist, um so behender ins fernste Ultraviolett hinaus, 

 je nither ihr die Fessel der lichtempfindlichen Platte riickt. 



Schon jetzt weist meine neue Platte jenseits 1852 ein Spec- 

 trumband auf, dass das gesamnite Wirkungsgebiet der Bromsil- 

 bergelatine um mehr als das dreifache an Lange iibertriflft, und 

 gleichwohl Ksst auch die letzte meiner Aufnahmen noch der 

 Hoffnung Raum, dass jenseits des Randes ibrer Platte noch pho- 

 tographisch wirksames Licht existirt. VorlSufig gehort aber 

 diese letzte Aufnahme, ohngeachtet solch' giinstiger Aussicht, 

 doch demjenigen Gebiete an, das ich gegenwSrtig als die Grenze 

 der kleinsten Lichtwellen bezeichnen muss. Die Photograpbie 

 des Nachbargebietes hiervon stosst zur Zeit, — aus Grunden, deren 

 Erorterung bier zu weit fiihren wtirde, — auf Hindernisse, die sich, 

 sofern es iiberhaupt moglich ist, nicht ohne grosse Anstrengung 

 warden beseitigen lassen. 



Fragt man nun nach dem Masse der kleinsten Lichtwelle 

 meiner Ultraviolettaufnahmen, dann muss ich leider bekennen, 

 dass mir im Augenblick eine beatimmte Antwort hierauf nicht 

 moglich ist. Wellenlangen lassen sich im luftleeren Raume, 

 an den meine Aufnahmen gebunden sind, nicht so leicht ermit- 

 teln wie in der Luft, und die geplanten Messungen der Wellen- 

 langen des aussersten Ultraviolett haben darum auch besonderer 

 Vorbereitungen bedurft. War es doch iiberhaupt zweifelhaft, 

 ob sich die iibliche Methode der Messung der Wellenlangen 

 auf den in Rede stehenden Lichtbereich werde anwenden lassen. 

 Meine Vorversuche hierzu gehen zur Zeit ihrem Abschluss ent- 

 gegen, und die mir vorliegenden Resultate berechtigen zu den 

 besten Hoffnungen. Unter solchen Umstanden kaun ich das 

 Mass der kleinsten Lichtwelle, die meine Aufnahmen aufweisen, 

 vorlaufig nur schatzungsweise und mit Vorbehalt nennen. Es 

 diirfte dieses Mass 1000 AE nicht iiberschreiten, ja eher kann es 

 um ein gutes Theil kiirzer sein. 



Der Wellenlange 1000 entspricht eine ausserordentlich hohe 

 Schwingungszahl des Lichtathers. Wahrend die brechbarsten 

 Sonnenstrahlen wenig mehr als 1000 Billionen Schwingungen in 

 der Seounde ausiiihren, schwingt ein Strahl von der Wellenlange 

 1000 in derselben Zeit dreibillionenmal. Mit Schwingungszahlen 

 so enormer Hohe hat der Spectroskopiker bisber noch nie zu rech- 

 nen gebabt, und gleichwohl liegt es nicht ausser dem Bereich der 

 Mogliobkeit, dass wir iiber kurz oder lang die Whkungen des 

 Lichtaetbers bis in die t'achste Nahe der Wellenlange Null verfol- 

 gen werden, wo der ungebeuren Anzahl seiner Schwingungen 

 kaum noch der Massstab des Endlichen gewachsen ist. 



THE SOUNDS OF R.^ 



BY ALEX. MELVILLE BELL. 



There seems to be special need for a better understanding of 

 the sounds of R. No element of speech is so variously pronounced 

 — in dialects and by individuals. The fundamental organic ac- 

 tion from which all the varieties are derived is a friotional emis- 

 sion of bi-eath or of voice between two surfaces in tbe breath 

 channel. Thus we may make an R in the throat, — producing the 

 effect which, when prolonged, is called a groan ; or in tbe guttural 

 passage, between the back of the tongue and the soft palate — a 

 mode which is dialectically common in many countries. A less, 

 definite variety is formed between the arched top of the tongue 

 and the roof of the mouth. This is common in the United States. 

 Another — and the normal English form of R — is produced be- 

 tween the point of the tongue and the upper gum. This is some- 

 times modified by inversion of the tongue within the palatal arch, 

 or by addition of guttural or of labial contraction. The point- 

 tongue R is also varied by advancement of the tongue to or be- 

 tween the teeth. In a common English affectation the seat of 

 R is transferred fron the tongue to the lips, so that R has the 

 sound of W. Of these varieties one may be characteristic of a 



^ Paper read before the Phonetic Section of the Modern Language Associa- 

 tion, December. 1891. 



dialect, another a mere individuality, but they are all effects 

 of only one organic action performed at different parts of the 

 mouth. 



Another series of R's results from a trilling or rattling organic 

 vibration instead of a mere friction of the breath or voice. Thus 

 a trill of the epiglottis is heard as one form of R; a trill of the 

 uvula is another and very common one; and a trill of the point of 

 the tongue is the regular form of R in North Britain and Ireland. 

 The Spanish R has a more prolonged rattle of the same kind. 

 The trill has often the effect of a syllable; as in Scotch and Irish, 

 where it converts the grammatical monosyllables world, harm, 

 mourn, etc., into the phonetic dissyllables wor-rld, har-rm, etc. 



These trills involve a strong pressure of breath and a harshness 

 of phonetic effect, in contrast to which is a form of R of simple 

 vowel quality, without friction or vibration; as in (a)ise and 

 (a)ound, for rise and round. 



A similar vocalic effect is also heard for R wherever it is not fol- 

 lowed by a vowel; as in here, care, fire, store, tour, are, war, 

 term, first, etc. The syllable-like quality of this sound is dis- 

 tinctly felt after the close vowel e, and less distinctly after open 

 vowels, because their mouth- cavities differ so little from that 

 of R. 



In Early English R was always trilled, as it continues to be in 

 Scotland, where most of the characteristics of Early English are 

 still prevalent. But in modern English the trills have been soft- 

 ened away wherever R follows a vowel, until little is left of the 

 R but its vowel quality. We are accustomed to the entire omis- 

 sion of R in negro speech, where do and sto are all that we hear 

 for door and store; but in educated utterance there is some pho- 

 netic effect left in R even where it is least manifest. Such deli- 

 cate shades of sound aie the distinguishing marks of refinement 

 in pronunciation, and they should be carefully preserved by 

 teachers and by writers on phonetics. 



In a book recently published in England the learner is taught 

 that R is silent in such words as farm, serve, lord, prayer, weird, 

 etc. Had the statement been that the sound of consonant-R is 

 not heard in these words it would have been correct, but the R is 

 certainly not "silent;" it has a phonetic effect of its own, soft 

 and vowel-like, but a quality wanting which the words would not 

 have their characteristic pronunciation. 



That there may be no mistake as to the teaching in the w^ork 

 referred to, the reader is specifically told that the words arms and 

 lord are exactly the same to the ear as the words alms and laud. 

 Now what is the sound of R which baffles the discrimination of 

 this writer? Let us magnify it, as in a microscope, by prolong- 

 ing the elementary sounds. First let us put "alms" and " laud" 

 under the microscope : — 



a ----1ms; lau d. 



Here there is no R; the vowel remains unchanged until stopped 

 sharply by the succeeding consonant. Now put "arms'' and 

 " lord " under the microscope : — 



a Ca)rms ; lo (a)rd. 



Here between tbe vowel and the m or d there is interposed a 

 gliding connective sound, so that the vowel is not stopped sharply 

 by tbe consonant, but its quality is gi-adually changed by a lift of 

 the tongue, verging towards but not quite reaching tbe position 

 for R. This is all the sound that R has, in modern English, be- 

 fore any consonant or when final in a word. But it is some- 

 thing more than nothing; and something that is essential to 

 the correct utterance of any word containing R before a con- 

 sonant. 



Among the sounds of R may be reckoned the influence of R 

 upon other sounds. The mouth- cavity for R being very large, any 

 closer vowel preceding R is, as it were, stretched at the point 

 of junction, so as to assimilate with R. Thus a pure B is with 

 difficulty pronounced before R; a pure a is never, in Anglican 

 speech, heard before R, but a is " stretched " to eh, as in air, 

 chair. So, too, o and oo before R have a more open than their 

 usual formation, as in old — ore; pool — poor. 



These widened sounds of o and oo are distinctly different from 

 the sound of aw ; yet in the book before referred to the words 

 shore and drawer are said to have the same vowel; and the words 



