The claim:
We can know apriori that whoever invented the zip invented the zip. That the person who invented the zip invented the zip is a contingent historical fact, however. He/she might instead have invented pork bellies, or died aged 2. Nevertheless, we know he/she didn't. We know that whoever invented the zip, invented the zip. So, we have apriori knowledge of a contingent historical fact.
note: this is typically presented under the rubric of proper names whose meaning is fixed using a definite description; however, when Evans wrote about this first, he was trying to do two things - 1) vindicate what he saw as Frege's descriptivist theory of the reference of proper names, and 2) show that there was such a thing as the contingent apriori; Proper names are only needed to make the first point - but are not at all necessary for making the second point. So it's easier to think about the contingent apriori without getting the name 'Julius' (defined as the inventor of the zip) involved.
The 2d analysis:
The primary (de dicto) intension of the sentence ‘Julius invented the zip’ is ‘the inventor of the zip invented the zip’, and hence is true in all possible worlds. The secondary (de re) intension of the sentence is ‘Will Whitworth invented the zip’ in our world, and it is true in our world. However, there are possible worlds where Tiny Tim invented the zip. And in those worlds, the secondary intension ‘Will Whitworth invented the zip’ is false.
The worry:
The main worry I can think of is that perhaps nobody uniquely invented the zip (quantum accident, team work, whatever). This was raised by Evans in his original (1977). In that case, we'll have to state our claim as a conditional - if someone invented the zip, the inventor of the zip did. But is this apriori knowledge of a contingent fact? It doesn't sound like it anymore - we don't know if anyone invented the zip, so what historical contingent fact do we know? We know a conditional that is true irrespective of history - in any possible world, if anyone uniquely invented the zip, the person who invented the zip did. This is true, but not contingently true - it's true even if nobody invented the zip. There is no historical fact that could change the truth value of that claim, which is true even in worlds with no zips; so it isn't contingent on any other truth or fact.
We might think that this is a false alarm, however, and that Evans' amendment in fact causes more trouble than it solves. I can have apriori knowledge on the basis of knowing other things - if I know it's raining, then I can know apriori that it's wet. Similarly, if I know someone invented the zip, then I know that the person who invented the zip invented the zip.
