In this third paper on sulfur species, we have conducted a survey of S
O+ (two transitions) and H2S (one transition) in our standard samples
of 11 cirrus cores and 27 Clemens-Barvainis translucent objects whose
structures and chemistry have been studied earlier in this series. S0(
+)((II1/2)-I-2, J = 3/2-1/2) is seen weakly in 12 objects, while H2S (
1(10)-1(01)) is detected quite strongly in 31 objects. These results a
re modeled in terms of our previous hydrostatic equilibrium and n simi
lar to r(-alpha) structures together with other chemical and physical
properties derived earlier. The typical H2S fractional abundance is la
rge, similar to 1 x 10(-8), and increases monotonically with increasin
g extinction in the 1.2-2.7 mag range (edge-to-center). Thus H2S displ
ays the same characteristic transition between diffuse and dense cloud
chemistry as do SO, SO2, CS, HCS+, HCO+, and other species studied in
this series. By contrast, the SO+ abundances are small, 1 x 10(-9), a
nd exhibit a marginal decrease with increasing extinction. The simple
ion-molecule network as used by Turner for sulfur chemistry includes t
he sulfur hydride species and predicts the observed parameters of SObut predicts an H2S abundance 2 orders of magnitude less than observed
. Of the 10 species presently analyzed in detail in the translucent co
res, H2S is only the second (along with H,CO) that fails to be explain
ed in detail by quiescent cloud ion-molecule chemistry. Various cataly
tic models of H2S on grains are discussed. Photocatalysis of H2S is fo
und capable of producing the observed abundances but only for sizable
accreted mantles. Other types of surface chemistry are also successful
but are close to the limits of possible efficiencies. We have detecte
d OCS and H2CS in one object, CB 17, with abundances of 1 x 10(-9) and
7 x 10(-9) respectively. Our ion-molecule model has been expanded to
include OCS and H2CS chemistry. We find that the model fits observed a
bundances within a factor of 3 for both species.