Remark on the nature of the spectrum of the lamé equation. A problem from transcendence theory