16#16 can be derived from simple arguments about stars on the Hayashi track. The potential energy given off by stars on this track is 17#17, while the stellar luminosity 18#18. Since the stellar luminosity is the change in energy per unit time, 19#19. Along the convective tracks, temperature is approximately constant. Removing all the constants one gets 20#20. Integrating leads to 21#21.

The Image Reduction and Analysis Facility is distributed by the National Optical Astronomy Observatories.

The usage of S/N in this context is not the same as in standard photometry. ``Noise'' has the same meaning, but here ``signal'' refers to the intrinsic strength of the modulation. If a star is measured to 2 accuracy and has at 10 modulation (0.2 magnitudes peak to peak), the S/N is 5.

The sensitivity limit varied from field to field on the basis on the brightest available field standard star. The mean noise level in the field standards was about 5.

Edwards et al. (1993) uses H-K=0.2 as their cut off for stars without disks. I use 0.4 as an upper limit because the errors in H-K can be 213#213 0.1.

It is still possible to plug the cord to the drive in upside-down. The operator should take care in placing this plug.

The accuracy of the results presented Tables gif and gif is often limited by the resolution of the grid used for our simulations. In many cases, the last digit tabulated should not be regarded as significant, so results which appear to be identical to four places despite differing input parameters may only have been the same to three places if a higher resolution code were used. However, given the scatter in the observational data, we do not believe that more precise modeling is warranted.

Scott J. Wolk
Mon Nov 25 15:41:03 EST 1996