We present HST STIS/G140M spectra of 15 extragalactic targets , which we combine with GHRS/G160M data to examine the statistical properties of the low- z Ly \alpha forest .
With STIS , we detect 109 Ly \alpha absorbers at significance level ( SL ) \geq 4 \sigma 
over 0.002 < z < 0.069 , with a total redshift pathlength \Delta z = 0.770 .
Our combined sample consists of 187 Ly \alpha absorbers with SL \geq 4 \sigma over \Delta z = 1.157 .
We evaluate the physical properties of these Ly \alpha absorbers and compare them to their high- z counterparts .
Using two different models for Ly \alpha forest absorbers , we determine that the warm , photoionized IGM contains 29 \pm 4 \% of the total baryon inventory at z = 0 ( assuming J _ { 0 } = 1.3 \times 10 ^ { -23 } ergs cm ^ { -2 } s ^ { -1 } Hz ^ { -1 } sr ^ { -1 } ) .
We derive the distribution in column density , N _ { HI } ^ { -1.65 \pm 0.07 } for 12.5 \leq \log { \left [ N _ { HI } ( { cm } ^ { -2 } ) \right ] } \leq 14.5 , breaking to a flatter slope above \log { \left [ N _ { HI } \right ] } \approx 14.5 .
As with the high equivalent width ( \mathcal { W } > 240 ~ { } { m } \mbox { \AA } ) absorbers , the number density of low- \mathcal { W } absorbers at z = 0 is well above the extrapolation of d \mathcal { N } /d z from z > 2 .
However , \log \left [ ( { d } \mathcal { N } / { d } z ) _ { z = 0 } \right ] = 1.40 \pm 0.08 for \mathcal { W } > 240 ~ { } { m } \mbox { \AA } is 25 \% below the value obtained by the HST QSO Key Project , a difference that may arise from line blending .
The slowing of the number density evolution of high- \mathcal { W } Ly \alpha clouds is not as great as previously measured , and the break to slower evolution may occur later than previously suggested ( z \sim 1.0 rather than 1.6 ) .
We find a 7.2 \sigma excess in the two-point correlation function ( TPCF ) of Ly \alpha absorbers for velocity separations \Delta v \leq 260 km s ^ { -1 } , which is exclusively due to the higher column density clouds .
From our previous result that higher column density Ly \alpha clouds cluster more strongly with galaxies , this TPCF suggests a physical difference between the higher and lower column density clouds in our sample .
The systematic error produced by cosmic variance on these results increases the total errors on derived quantities by \sim \sqrt { 2 } .