SOOM, _vb._ to swim. Not Dan. _sömme_, but loss of _w_ before _oo_, cp. the two Norse forms _svömma_ and _symma_. Cp. _soote_, the last word in the first line of the Prologue to Chaucer's Canterbury Tales. TEEM, _vb._ to empty. It is not necessary to derive this from Norse _tömme_, "to empty." There is an O.E. _tōm_ from which the Sco. adj. _toom_ probably comes. _Toom_ is also a verb in Sco. _Teem_ is simply this same word by characteristic Sco. change of _o_ to _e_. (See §17.) This also explains the length of the vowel. TRAK, _vb._ to pull, not necessarily Norse _trekka_, cp. the L.G. _trekken_. WID, _sb._ wood. Not O.N. _viðr_ nor Dan. _ved_. The vowel is against it in both cases. But just as above _toom_ becomes _teem_, so _wood_ > _wid_, cp. Sco. _guid_, "good," _pit_, "put," etc. (See §17.) Hence also the shortness of the vowel in _wid_. WERE, _sb._ spring, cp. Latin _ver_. _Var_, _vaar_ in Scand. does not account for the _e_ in the Sco. word. YIRD, _sb._ earth. Not from Dan. _jord_. See next word. YIRTH, _sb._ earth, an inorganic _y_ (see §18). Not from O.N. _jörð_. For _d_ in _yird_ see §19. 24. LOANWORD TESTS. I have adopted the following tests of form, meaning and distribution in determining the Scand. source of loanwords: 1. The diphthong _ou_, _ow_ corresponding to O.N. _ou_, O.E. _ea_. 2. _Ai_, _ay_ corresponding to O.N. _æi_, O.E. _ā_ as far as such words can be determined from modern dialects according to §20. 3. The spirant _th_ corresponding to O.N. _ð_, and O.E. _d_. 4. Consonantal assimilation of _nk_ to _kk_, _mb_ to _bb_, _mp_ to _pp_, _ðl_ to _ll_, _zd_ and _rd_ to _dd_, corresponding to similar assimilation in Scand. 5. Other consonantal and inflexional forms that are Scand., as opposed to O. Nhb. _d_ for Scand. _d_, O.E. _ð_ excluded, see §§19 and 23. 6. A word that is used in a sense distinctively Scand., as opposed to Eng. or L.G., is to be regarded as a loanword. 7. The distribution of a word in South England diall., or in O.F., O.S. or M.L.G., indicates that the word is not a Scand. loanword. 8. On the other hand, if a word occurs exclusively in Scand. settlements in England and Scotland, it is to be regarded as due to Scand. influence in Scotch in spite of L.G. parallels. 9. The presence of a word in O.E. excludes Scand. influence, except in cases where the O.E. word has been shown to be a loanword. See Steenstrup and Kluge. 25. REMARKS ON THE TEXTS. The following dates it may be well to remember: