that is.. a stretch. so they didnt actually look at sunlight hours they simply assumed the probability from the distribution of the data?? both hypotheses are plausible (sunlight vs magnetic fields) but this reads like data fitting rather than treating each possibility equally blind. where was it published?