There are several explanations for these X--ray quiet PMS stars. To examine the explanations, it is proper to divide the PMS stars into three groups: those bluer than R-I = 0.7, those redder than the X--ray sources (R-I > 1.25) and those with colors similar to the X--ray sources. For the first group, contamination by background stars is a problem. The peak in the distribution of background stars appears at R-I of 0.5. At this color, it is possible that dozens background stars are occupying the same region of the H-R diagram as PMS stars. However, since this group is also the brightest, we were able to obtain spectra for many of these objects.
A total of 26 of the non--X--ray detected PMS candidates
had spectra taken. They were part of the control sample of non--X--ray
emitting stars which was drawn drawn for the Guide Star Catalog
(Lasker et al.\
1990).
The results are given in in Table 18.
Strong (EW > 0.200 Å) Li I absorption features were found in
absorption in 19 of them. Of the 19 confirmed PMS
stars, nine were near 
Orionis. Two of these had strong
H
emission lines associated with cTTs. A preliminary result
is that 73% 
15% of the PMS candidates
identified solely through their optical colors, are indeed PMS.
As stars get redder, there is a
clear gap between the background sources and the PMS candidates.
The bifurcation between the background group and the X--rays sources
is almost a magnitude for stars redder the R-I = 0.7 and increases in
the redward direction. These stars were too dim to be observed with
the spectroscopic survey. To estimate contamination for the stars
near 
Orionis, I divided the color--magnitude diagram of
from R-I of 0.7 to 1.25 into 4 bins. The bins were defined by lines
running parallel to the reddening vector, and passing through
R-I = 0.7 and V magnitudes of 14 through 18. (See
6 for more details.) The group of stars
below the bottom line was ignored since this bin is incomplete.
In the lowest bin there are 63 stars, in the bin above that there are
24, in the next bin (the once that most closely fills the gap between
the background sources and the PMS stars) there are 10 sources.
A fit to these data projects 3--4 background sources to be found in
the top bin. If I had accounted for the incompleteness in the
bottom two bins, this number would be even smaller. Excluding X--ray
sources, there are a total of 22 sources in the top bin. Based on the
above argument, I expect about 85% 
10% of these to be PMS stars.
The existence non--X--ray detected PMS stars implies one of two conclusions.
The first possibility is there are X--ray quiet PMS stars.
It is possible that these stars are relatively slow rotators, so that
the dynamo for generating the activity is not as powerful. The
second possibility is that the activity rates in stars change with time.
Figure 9 shows the X--ray count rate of all but
the brightest detected X-ray sources near 
Orionis. The bottom 12
sources in the figure are within 3 
of the detection
limit.
Given that the observed X--ray variability of PMS stars in Taurus is up to a factor of five (Walter et al.\ 1988), many of the sources observed could be missing in later observations and many stars not observed here may be seen in future observations. The X--ray flux changes may be secular in nature or could indicate stellar activity cycles in these stars, similar to those of the Sun and older stars (cf. Dorren 1995). However, Walter et al.\ (1988) note that of Taurus--Auriga stars observed in X--rays more than once, 25% of them have significant source variability which led to non--detections during at least one observation. Between V=14 and V=16, I find 25 stars which are X--ray sources and estimate 19 non--X--ray PMS stars.
Based on their colors the non--X--ray detected stars are PMS.
The group of stars which are redder than the X--ray sources, R-I 
,
is probably composed
of stars similar to the X--ray sources, except that their smaller mass
leads to a smaller surface area. Assuming a constant X--ray flux per unit
surface area, the lower mass stars are X--ray dim.
Consequently, they fall below the ROSAT detection limit.
