120 WISCONSIN ACADEMY SCIENCES, ARTS, AND LETTERS. 



terite from the assimilated Jujcgan, we would have hycde> hycte> 

 hyhte, which could not have been distinguished from hi/hfe, the 

 preterite of Jiylitan^ to hope. It is evident that liogde with the 

 Rilck-umlaut is a later form than hygde, if not, there is no reason 

 why we should not have Jiohte. The cognate hogjan (Class II.), 

 pret. Jiogode, may have been the cause, that hygde adopted Rtlck- 

 umlaut. In thyncan the y is Umlaut of u ; in the preterite Rlick- 

 umlaut takes place; ththte <thilcte <thuncte. In ivyrcan the y is 

 Umlaut of 0, weakened from u; as y and i are interchanged, we 

 find sometimes wircan, which has the breaking weorcan, but wircan 

 and iveorcan are bad spelling, for while y might always replace /, 

 the reverse could happen only after it had been forgotten that y 

 was the Umlaut of u (o). In the preterite the original vowel reap- 

 pears regularly, as worcte> ivorhte ; the o is therefore not the effect 

 of the A, as Dr. March assumes in the article above mentioned, (p. 

 112, 3); for the original y=i would have the breaking eo. The 

 lecgan and secgan arise through assimilation and umlaut from 

 lagjan and sagjan. As the derivative J ((/), Anglo-Saxon e, disap- 

 pears in the preterite of stems long by nature or position, which 

 causes Rilck-umlaut, — the'preterite of these verbs is '^lagede,*sagede 

 ^laegde, saegde, contra., laede, saede; ae is the regular weakening 

 of a. The e in segede, legde is bad spelling for ae. 



In Old Norse we observe similar euphonic changes, e. g. soekja 

 from soJi-ja^ oe {ae) being umlaut of o, forms the preterite with 

 Riick-umlaut sotta. We should expect sokta, tor kt never assimi- 

 lates into it, as Helfenstein says; but as k represents an original h, 

 which with t assimilates into tt, the form sotta is entirely regular. 

 In Old Sivedish we find sold by the side of sott. In yrkja y is 

 the umlaut of o (a); its ^xeieriiQ is orta 2indi orkta(varkta). Of 

 three consonants one is sometimes dropped, cf. mart for margt; 

 morni for morgni, apni for aptni; thykkja or thykja, kk==nk, pret. 

 thotta from th6hta< tkok(n)hta; thekkja, e is the ^umlaut of a, 

 pret. tJidtta from thdhta<thok(n)hta. 



2. — THE USE OF THE INFINITIVE OF MODAL VERBS INSTEAD OF THE 

 PRETERITE PARTICIPLE IN NEW HIGH GERMAN. 



Fred. Miinch, a well known German-American writer, advanced 

 lately the opinion that it was a blunder to say, "ich habe es tun 

 kOnnen," instead of ''ich habe es tun gekonnt." In reference to 



