Your mistake is in assuming that because we see varves being put down annually that all varves are put down that way.
In this particular case, this is not an assumption.
(Thanks to ImageShack for
Free Image Hosting.)
The data in the image above are radiocarbon dates for layers in several different phenomena. As you can see, they all match almost perfectly. The data includes very different phenomena, such as tree rings, lake varves, and mineral deposits in a cave.
The data make perfect sense if one assumes that these are annual features as they appear to be. If these are not annual, then one has to wonder exactly how these very different phenomena manage to track one another perfectly like this.
This is what is called consilience of data. We have different, completely independent lines of evidence that the varves in, say Lake Suigetsu (which is one of the data sets above) are annual. If they weren't, we can't explain why they track so perfectly other, completely independent data that are also usually annual.
(I can't find the original graphic -- I got this one from
TheologyWeb.
"The guilty one is not he who commits the sin, but the one who causes the darkness."
Clearly, he had his own strange way of judging things. I suspect that he acquired it from the Gospels. -- Victor Hugo