AUTOPHOV. 



AUXILIARY VERBS. 



on UM pianoforte, w*r also exhibited 



years ago in 



mystery. A bo; 

 r solated only 10 



A boy was * 





the figure. The great 



tj existed only so'long a. it was imagined that the player was 

 UM ftfnn; nenrth.)eas, taw machinery by which the hands 

 julated must hav* been ingvniooa. 

 Tk riissinn for making automate has not yet quite jiossed away. A 

 rwrt example was Mr. Facer's Buptonia. It consisted of a draped 

 boa* awl waxm-taced figure, which articulated language with a certain 

 degree of intelligibility. The sounds were produced by pressing on 

 sixteen hsya. A small pair of bellow* was worked with the nozzle in 

 the back part of the bead of the figure ; and in the head were various 

 arrangements of India-rubber and other materials, calculated to yield 



* particular sound in each part or section. When the exhibitor wuhed 

 to produce sentence or word, he first mentally divided it into as 



rto as tfcers arc actually distinct sounds not necessarily 

 ; with the syllables or the single letters ; since the various 

 >hic systems far more correctly represent distinct sounds, 

 aving determined the first word, the operator pressed his finger on a 

 particular key, which admitted a blast of air to a particular compart- 

 ment, in which the mechanism was of the kind to produce the sound 

 required. Other keys ware similarly pressed, until all the required 

 sounds of the word or sentence were produced. The sounds were 

 sufficiently analogous to those of the human voice to convey the 

 monini: intended, but they had an unpleasant effect on the ear. By a 

 modification of the action, whispering was imitated. 



A remarkable machine was the Automaton Latin Yernfer, intro- 

 duced in 1846, by Mr. John Clark, of Bridgewater, after a labour of 

 thirteen years. At the first thought such an invention seems inex- 

 plicable, owing to the mental character of the process; but a little 

 inquiry shows that it is only a system of permutations, such as a 

 machine on easily be made to produce. The specimens given in the 

 ' Athensnim ' and other public journals at the time are all I -it in 

 hexameters, and moreover have all the some grammatical formula and 

 scansion, in respect to dactyls and spondees. The following nine speci- 

 mens were given, each complete in iteelf as an huxametric line, but 

 having no connection with the others ; 



1. Horrida sponsa reis promittunt tempera densa. 



2. Sontia tola bonis cansabunt agmina crebra. 

 8. Bellica vote modis promulgant crimina fusca. 



4. Aspera pila patet depromnnt pnclia quaxlam. 



5. Effera sponsa fere confirmant vincula nequam. 



6. Barbara tela reis prsemonstrant nubila dura. 



7. Horrida vota bonis progignunt jurgia crebra. 



8. Sontia castra inodis prositant somnia funca. 



9. Trucida regna quidem conquirunt opera cam. 



The exterior of the machine which composed these lines resembled 

 in size and shape a small bureau bookcase; in the frontispiece of 

 which, through an aperture, the verses appeared in succession as they 

 were composed. Mr. Clark, in a communication to the ' Athenccum ' 

 (No. 923) made the following observations on his machine : " The 

 machine is neither more nor less than a practical illustration of the 

 law of evolution. The machine contains letters in alpha- 

 betical arrangement ; out of these, through the medium of numbers, 

 tendered tangible by being expressed by indentures on wheel-work, 

 the instrument selects such as ore requisite to form the verse con- 

 ceived : the components of words suited to form hexameters being 

 alone previously calculated, the harmonious combination of which will 

 be found to be practically interminable." 



Mr. A. J. Cooler, in the same journal, pointed out the existence of 



forgotten pamphlet, a century and a-half old, in which the author 

 bowed how, from a table given, a person might produce millions of 

 hexameter lines. But these wore produced by accumulations of words; 

 whereas Mr. Clark's machine, if we rightly understand his description, 

 actually builds up the lines letter by letter. 



In 1854 a strange automatic group was exhibited in London. A 

 figure it presenting a monkey held a violin, moved a bow across the 

 strings, pressed the fingers alternately on them, and clapped its jaws 

 together in token of satisfaction. A hare browsed, or seemed to 

 browse, at a cabbage. A goat uttered on audible cry. A doll-child, 

 that had been quietly reposing in its cradle, woke up uneasily, and 

 screamed aloud for iU " Pa" and " Ma." The doll was a failure, in 

 regard of resemblance to a living child ; but the monkey, hare, and 

 goat were cleverly constructed, the internal machinery being covered 

 with the real skin of the animals ; the form of each animal being well 

 imitated by padding within the skin ; and the slight movements of 

 each animal, such as the twinkle of the eye and the twitch of the toil, 

 being reproduced with great exactness. 



It affords matter for regret that so much ingenuity should )H,> 

 expended on the production of automata leading to no useful results. 

 There are, however, many machines for calculating, numbering, 

 registering, stamping, pageing, Ac. which illustrate the application <,t 

 automatic action to useful purposes. Some of these will be noticed 

 under CALCCLATIKO MAUIIXES. 



A0TOPHON, a kind f barrel-organ, the tunes of which are pro- 

 duced by meant of perforated sheets of mill-board, instead of pins or 

 studs, as in the ordinary barrel-organs. 



AUXILIARY VERBS at* distinguish** from other verbs in the 

 following way. Verbs express the notions of action : auxiliary verbs, 

 though they originally expressed notions of action, only express 

 nfa/iMM of adiM when considered as auxiliary verbs, and are accord- 

 ingly employed, in connection with other verbs, to give to thorn certain 

 relations called by grammarians tense, mood, and voice. The modern 

 languages of Europe, and our own more particularly, abound in such 

 forms ; but they are likewise found in the languages of Greece and 

 Rome, sometimes altogether undisguised, more commonly so completely 

 blended with the main verb a* to pass for a mere arbitrary suffix, 

 which the grammarian does not attempt to explain. It is in the very 

 nature of a particle which plays a secondary part, that it should not 

 occupy too large a share of the attention ; and thus those verbs which 

 in course of time are used as auxiliaries, though originally u significant 

 as any other verbs, lose something of their distinctive character ; so 

 that if the fuller form happen to disappear from a language, the 

 corrupted auxiliary presents anomalies which it is not easy for the 

 philologist to explain. This difficulty is increased by the circumstance, 

 that verbs used as auxiliaries generally throw off much of the distinctive 

 meaning which they originally possessed. 



Among the auxiliaries, the most important is the substantive verb 

 signifying to be ; and, as might be expected, no word has passed through 

 more variations of form. Grimm and other grammarians, indeed, have laid 

 down, that there are three, or even more, distinct roots combined in the 

 conjugation of thU verb. But when allowance is mode for the known 

 changes that take place in the letters of the alphabet, there will appear, 

 we think, some reason for supposing that all the varying forms of this 

 verb are derived from a common origin. 



As the ultimate form from which all the rest appear to us to have 

 flowed, we will propose the root IK* ; and we are inclined to assign to 

 this root, as its primary meaning, the notion of taliny. Such n form 

 appears in the Latin retear (pronounced wescor), / eat, and in the 

 German tea-en, to be. The initial tr, it is well known, sometimes 

 assumes the form of g, and hence we have ge-get+tn, eaten. Again, as 

 the German verb leten to read forms a past tense, er lot be read, so 

 veten accounts for the form vat, common, as the past tense, to English 

 and German. Still more commonly the re is altogether dropped, 

 and then we have the root et, which is the basis of the Greek sub- 

 stantive verb a-mi (the original form), et-ti, et-ti (still existing in this 

 form in the Lithuanian language), of the old Latin verb et-vm, a, t-t, 

 et-vmia, et-tit, ei-unt, et-to, ta-sr, and with a slight variation of the 

 Sanskrit ai-mi, Ac. With the same form of the Latin we may connect 

 ti-t, he eatt, tt-te, to eat,et-ea, a-culentiu, ic., and the German ea-rx, to 

 eat. Aftertbe word has thus been stripped of its initial consonant, tho 

 short vowel also was apt to disappear, at least in the longer forms. 

 Thus from the old Latin forms etum, emnt, aim, Ac., there arose the 

 shorter forms mm, tunt, tim, &o. ; prae-tt-ent, ab-tt-nent, were reduced to 

 praetnu, abitiu ; and in German we find trin, to be, find, Ifiey are, in 

 place of a-ein, a-t'nrf. 



In the second place, the consonant i interchanges with the lett.-r r, 

 so that vert exists by the side of inu, and art, are, with t*. Thus in 

 the Latin too we have er-am, er-o. where more regular forms would 

 have been et-am, ei-o. Again, the same letter i is interchangeable with 

 the dentals I, </. Hence, while the Germans have en-en, irk au, the 

 English express the same notions by to eat, I ate; and the Latin 

 tongue uses indifferently ed-il or et-t, he eatt, ed-ere or tue, to eat. 



The form be is evidently the parent of the German 6m, 7am, lint, 

 tfum art, and of the English be-iny and be-rn. With the short vowel 

 changed, it appears in the Lithuanian fru, as bu-ti, to be, buimu, I hare 

 been; and as 6 in English generally corresponds to /in Latin, we must 

 claim the Latin fu-ri or fu-i,fu-am, fu-turut, 4c. That these forms 

 are all related among themselves is generally allowed ; Imt the question 

 now proposed is, whether they are not also radically connected with 

 the root vet. If it could be shown that the root be ever existed with an 

 .' at the end, it would no longer be thought a violent step to suppose a 

 connection between bet and vet, more especially when we find th.- 4 

 already half way towards a w in fui. Now, a strong presumption that 

 the root be had a sibilant, arises from the old German form bir-vmet, we 

 are, compared with war-umet, ve were, in the same language (see 

 Grimm's ' Deutsche Grommatik '). In these words the suffix, which 

 denotes the plural pronoun, cannot claim more than four letters nmrt, 

 thus agreeing very precisely with the Doric Greek suffix omet, the 

 Latin um ut, and the Lithuanian a me of the gome power. Tin- radical 

 parts then are Wr and irar ,- and as we know the latter to be connected 

 with the form wot, there is no slight suspicion that Wr implies an 

 parly form, bit. If the Greeks lost the t in many of their forms derived 

 from the short root et, as they did, and if we ourselves have dropped 

 it from am, we can scarcely bo surprised at its disappearance from the 

 longer form bet or bit. The notion that the roots bet and vet are con- 

 nected, U confirmed again by the other forms in these languages, 

 which represent the idea of eating. In Greek, we find bo-ico, bo-tot, 

 io-ra, in Latin patco, pateor, as well as rtteor. 



The use of this auxiliary in the passive, both in ancient and modern 

 languages, is familiar to all; but it has been less carefully oK 

 that it is likewise employed in the perfect tenses of the active voice, at 

 least in the Latin language. Amar-eram, amar-ero, amat-ittem, amar-ittr, 

 evidently contain the forms cram, era, raem, eue; and in the perfect 

 subjunctive, an older form, amaretim, may be inferred from th 



