in the modification the prototype has been subjected to and in spite of this morphological element, therefore, these ellipses belong principally to etymological phonology.

Thus in our opinion the most natural and appropriate division of the elliptical words originating in the spoken language should be based on the classification they present with respect to the three principal parts of etymological grammar. Consequently we institute the following division:

1) elliptical words belonging to etymological phonology.

2) elliptical words belonging to etymological semology.

3) elliptical words belonging to etymological morphology.

If we turn especially to words illustrating the first-mentioned category, we will find it appropriate to divide them into two dif ferent classes, according to their semological nature, viz. elliptical personal names and elliptical words that are not personal names. To the former category it is advisable to assign also ellipsis of such common substantives, designating living beings, as have been abbreviated when occasionally used in the same function as proper names. Such words are the names of relationship, e. g. sis for sister, unkie (nunks) for uncle, pap for papa, titles such as doc for doctor, miss for mistress, nicknames such as Chim for chimpanzee, personifications such as Giffy for giraff. Strych for strychnine. The same should be done with proper names used to designate living beings, e. g. Atthey for Atheneum as name of a horse, Lympy for Olympus as name of a dog, and also with names of mythological personalities, e. g. Nep for Neptune, Clotty for Clotho. All these denominations may be termed equivalents of personal names.

It is the elliptical phenomenon as revealed in English personal names and their equivalents that we will try to examine first.

CHAPTER II.

Ellipsis of personal names and their
equivalents.

Although the category of elliptical words we are going to examine here should principally be dealt with under the study of etymological phonology, yet the elliptical phenomenon as manifested in them also presents qualities falling within the range of the morphological and the semological point of view. Our investigation, therefore, will naturally be divided into three principal sections, according as we are concerned with the phonological, morphological, or semological point of view. In connection with the last, the causes of the curtailment will be examined, since the semological change brought about through ellipsis is chiefly conditioned by them.

But before we begin our examination of the elliptical phenomenon from these points of view, it will, for practical reasons, be advisable to present the material on which we propose to base our research. As the chief factor for the division of the material we will take the locality of the ellipsis, this being the point of view that affords the most important and conspicuous kinds of elliptical abbreviation. As a second basis of division it is appropriate to take the semological nature of the elliptical words. Consequently we shall make a distinction between Christian names and surnames, and include under one category all the names we have termed their equivalents.

Concerning the further arrangement of the words belonging to the categories thus distinguished, the following principles have been adopted. All pet-forms derived from the same prototype, but differing phonologically from each other, irrespective of the locality of the ellipsis, have been given separately as special forms, e. g. Dun(nie): Dunc<Duncan, Gatty: Gerty <Gertrude, Ned: Ed< Edward. The same has been done with pet-forms that are phonologi

cally identical, but belong to different prototypes, e. g. Milly< Mildred, Milly<Millicent, and Milly<Amelia. As the same prototype we have considered such doublets as Alexander: Alexandry, Angela: Angelina, Clemency: Clementina, Feremy: Jeremiah, Lettice: Letitia, etc. On the other hand, merely orthographic variants of the same elliptical word will be disregarded, so that such orthographic doublets as Gerty: Girty <Gertrude, Geoff: Jeff <Geoffry, Jeffry will not be given separately. The addition or the omission of the hypochoristic suffix has been stated only when we can give a reference for it, although the one or the other certainly occurs in many other cases. The orthographic form of the suffix y, -ie, -ey, that often varies even in the same word, has been indicated in the form it occurs in the references.

Occasionally pet-forms are met with that have the appearance of being ellipses of their corresponding full names, but which in reality have other origin. Such cases are presented, for instance by Abby when pet-form for Arabella, by Ada when pet-name for Adelaide, etc. These cases of pseudo-ellipsis must not be overlooked in an investigation of the elliptical phenomenon. In our collection of material they have provisionally been assigned to those elliptical categories to which they in appearance belong.

In most cases it has been necessary to prove the existence of the ellipsis through quotations from literature. These are also required from a semological point of view, since by showing under what circumstances the ellipsis occurs, its semological nature is indicated1.

In answering the question whether the preserved phonem undergoes any phonological change by the ellipsis, it is necessary to know the pronunciation of the pet-form and its prototype. The system we have adopted to indicate the pronunciation is SWEET'S phonetic transcription, as given in his »Primer of Spoken English» (Oxford, 1900).

As regards the pronunciation of the unstressed vowels, phoneticians hold different opinions about what is normal in educated Southern English, which is the pronunciation we have tried to indicate here. (Cf. STORM, Engl, Phil., i 420). Especially in this particular, therefore, we cannot claim to have invariably succeeded in indicating the normal phonetic form, the more so, as by adopting Sweet's phonetic transcription we have been forced also to adopt to a certain extent his radical view about the pronunciation of the unstressed syllables. Since the dictionaries generally give incomplete information about the prononciation of personal names, it has in most cases been necessary to ask the opinion of English-speaking people.

I. Final ellipsis.

A. Christian names.

1 CHARLES P. G. SCOTT writes (Trans. of the Amer. Phil. Assoc., xxiii (1892), p. 296): "The dictionaries say that Agnes is pronounced Ag'néz (ag'nis), implying that it is a Latin word pronounced according to the so-called "English method" of pronouncing Latin; but Agnes, as an English name, is, like Moses, Latin only in seeming, and is pronounced in English fashion, Ag'nes (ag'ness), in termination like Mo'ses (mō'zess). All the girls I ever knew named Agnes wer cald Ag'ness. The pronunciation Ag'nez is one of the pernicious results of book-learning."