>Hmmm....I hadn't thought about that....In testing it is(!).
>
>A 1 to 1000000 loop is faster that 1000000 to 1. 3.30 seconds vs. 3.35....

I got inconsistent results with the same range. Sometimes the first was faster, sometimes the second. That why I said that I was mistaken. For a moment I thought that the loop difference may be the reason behind the other test, but resukts their were always consistent.

>
>>>
>>Going from the first character to the last in the loop rather than the last in the first. I thought, mistakenly as it turns out, that incrementing the loop positively (from 1 to n) would be faster than decrementing it (from n to 1). No ig deal.:-)