[annotated] HST WFPC2 images were taken of three LSB dwarf elliptical galaxies in the Virgo cluster. The intent of the observations was to determine the small scale structure in these enigmatic galaxies, and to attempt to learn something about the nature of their giant branch through the detection of luminosity fluctuations. In two of the three studied galaxies, V7L3 and V1L4, the luminosity fluctuations were unambiguously detected, yielding a density of 2 - 10 red giants/pixel. Using the observed B-V and V-I colors as a constraint, we could find no model that would reproduce the observed fluctuation signal and blue colors if there was a significant population of M-giants in these systems. The third system, V2L8, did not have a detectable fluctuation signal which possibly implies it is not in the Virgo cluster. Interestingly, this system is highly nucleated. Our observations have resolved this nucleus and if V2L8 is in Virgo, then we have discovered what is likely the smallest bulge measured to date, having an effective radius of only 50 pc. This bulge is quite red (as red as giant ellipticals) and its entirely possible that this nucleated dE galaxy, in fact, is a very large galaxy located in the background. As such, it is highly reminiscent of the manner in which Malin-1 was discovered. Finally, we find no evidence for small scale clumping of stars in any of the studied systems at this much improved spatial resolution. This implies these systems are dynamically well-relaxed and that the physical cause of their observed low surface brightnesses is their low density. When imaged at the high spatial resolution of the WFPC2 (~6 pc per pixel), the galaxies are easy to look right through without evening knowing they are present in the very middle of the WFPC2 frame. They appear only as elevated ``sky noise''.