In his review James Clackson correctly points out that I underestimated the role of the root aorist in the formation of the Latin perfect system. I said on pg. 412 that "Latin does not continue any clear traces of the root aorist". In the footnote thereto I mentioned the possible explanation of the lengthened grade of vēnī as generalized from 1st sg. *gwēm < *gwem-m and 2nd sg. *gwēn < *gwem-s. It's true that no personal ending of the perfect system can be directly traced to a root aorist, but there are a number of perfect forms that I failed to mention that most plausibly continue root aorists. The best cases are:
1. OL fūī ' I was' matches Ved. ábhūt.