THE PERIOD-SHIFT EFFECT IN OOSTERHOFF TYPE-II GLOBULAR-CLUSTERS WITH BLUE HORIZONTAL-BRANCH MORPHOLOGIES

It is shown, on the basis of data available in the literature, that th e mean periods of RR Lyrae variables in Oosterhoff type II (OoII) gala ctic globular clusters are not significantly correlated with the HB mo rphology, as represented by the parameter (B - R)/(B + V + R), for obj ects with blue horizontal-branch (HB) morphology types [(B - R)/(B + V + R)>0.65]. This result is compared with detailed synthetic HB model predictions based upon the evolutionary tracks of Lee & Demarque [ApJS , 73, 709 (1990)]. The predicted period shifts are found to be strongl y correlated with (B - R)/(B + V + R), as a consequence of the effect of evolution away from the zero-age HB becoming progressively more imp ortant as the latter quantity approaches + 1.0. The slopes of the pred icted and observed [log P]-(B - R)/(B + V + R) relations are found to differ at confidence levels greater than or similar to 88%. We argue t hat these results are likely not a spurious consequence of statistical fluctuations affecting the clusters with the bluest HB types, nor of our choices of metallicities Z, HB morphology parameters (B - R)/(B V + R), and mass dispersions on the HB sigma(M), for the adopted sampl e of 8 OoII clusters with statistically significant numbers of RR Lyra e variables (N(RR) greater-than-or-equal-to II and/or N(ab) greater-th an-or-equal-to 7). We show that reconciliation between the models and the observations cannot be achieved without dramatic changes in the in stability strip topology and/or period-mean density relation, and thus suggest that either the adopted evolutionary tracks are inadequate-we find evidence that the Castellani et al. [ApJS, 76, 911 (1991)] track s already lead to a significantly better, though not fully satisfactor y, agreement between the models and the observation-or other parameter s vary among our sample of clusters besides age and/or mass loss on th e red-giant branch, in which case a strong impact would be expected up on the interpretation of the second-parameter and Oosterhoff dichotomy problems. Thus, our analysis clearly represents the confirmation of t he period-shift effect for OoII clusters, in a manner which is complet ely insensitive to uncertainties in the reddening and temperature dete rminations, and may also constitute evidence against its interpretatio n on a purely evolutionary basis.