I9IS.] PRINCE— PRONOUNS AND VERBS OF SUMERIAN. 43 



the grain which hath drunk no water in its bed ; 



sugur edin-na pa nu-slg-ga-nm 



kim-mat-su ina ci-e-ri ar-ta la ih-nu-u 



whose bud in the field no shoot has borne ; 



Gis-A-AM ifto(RAT)-Ma ha-nu-su{g)-ga-mu 



il-daq-qu sa ina ra-ti-su la i-ri-sii 



the sprout which in its water-ditch is not planted ; 



Gis-A-AM ur-ra ba-ab-bu-ra-mu 



{il-daq-qu) sa is-da-nu-us in-na-as-xu 



the sprout whose roots have been torn away ; 



gii sar-sar-ra a nu-nag-a-mu 



qii-ii sa ina mu-sa-ri-e me-e la is-tu-u 



the vegetation which in its bed has drunk no water ; 

 A similar construction to the above is undoubtedly that in 

 ASKT., p. 122, i6: eri-zu-ka dg-gig-gd ak-a-mu^ana ar-di-ki sa 

 ma-ru-us-tum ep-su, 'unto thy servant (fern.) who has (lit. 

 * makes ' = afe) sickness.' 



It is perfectly evident from the above examples that we have 

 here a purely relative -mu used with participles. This is probably 

 identical in derivation with the demonstrative mu- in the regular 

 relative pronoun mu-lit and also with the common wi(-prefix of 

 verbs. It is quite possible that this relative -mu was used to indi- 

 cate all three persons, like the wM-prefix in verbs. 



What then are we to conclude as to the pronominal use with the 

 Sumerian verb? Is it possible to imagine a verbal system with 

 no fixed method of expressing the pronouns? The existence of the 

 practically fixed second personal value of the infix -ra- and of 

 the very common use of ga- as a first person would lead us to 

 suppose that the verbal prefixes were really not indeterminate pro- 

 nominally, even though Delitzsch lays down the rule that there is no 

 second personal conjugation in Sumerian (p. 102). 



The existence of third personal elements has long been recog- 

 nized. The difficulty lies in the apparently indiscriminate use of 

 many verbal prefixes for all three persons and the fixation of this 

 usage by Poebel's undoubtedly valuable equations. The question 

 now is whether Poebel is right in supposing that there underlies in 

 €very case of a first personal usage the '-vowel, i. e., that mu-, ' I ' 



