the ip that was given of 18.104.22.168 when looked up does show as Google
not sure if that ip range of their's is used for their bots or if it might be used for their own staff, possibly a human at google may have been browsing or checking a site. could have been checking for cloaking via user agent.
Not knowing info about the site where the log entry was from (that is, if the site was written in non-english) a translator might have been legitimately been used.
Simialar types of these visits happen to me frequently.
As most are aware I have the majority of RIPE denied.
Some visitors will attempt to access a page and after the resulting 403 will come back immediately on the gooogle translator. Unfortuantely that gets 403'd as well.
Jim's prospective on the google translator is the most logical that I'm able to recall. He feels that is a visitor is intersted enough to use the google traslator (most software translators are pitiful and will be until the technology changes) than it's he desire to allow their visits.
Personaly, I'm just not willing to read or speak other languages so that I'm capable of determining if my pages have been duplicated. Hell! I have a hard enough time with the English visitors.