1: \begin{abstract}

2:   We identified 24 \ion{Si}{4} absorption systems with $z \lesssim 1$

3:   from a blind survey of 49 low-redshift quasars with archival {\it

4:     Hubble Space Telescope} ultraviolet spectra. We relied solely on

5:   the characteristic wavelength separation of the doublet to

6:   automatically detect candidates. After visual inspection, we defined

7:   a sample of 20 definite (group G = 1) and 4 ``highly-likely'' (G =

8:   2) doublets with rest equivalent widths \EWr\ for both lines

9:   detected at $\ge 3\sigEWr$. The absorber line density of the G = 1

10:   doublets was $\dNSiIVdX = 1.4^{+0.4}_{-0.3}$ for $\logSiIV >

11:   12.9$. The best-fit power law to the G = 1 frequency distribution of

12:   column densities \ff{\NSiIV}\ had normalization $k =

13:   (1.2^{+0.5}_{-0.4}) \times 10^{-14}\cm{2}$ and slope $\aff{N} =

14:   -1.6^{+0.3}_{-0.3}$. Using the power-law model of \ff{\NSiIV}, we

15:   measured the \Sithr\ mass density relative to the critical density:

16:   $\OmSiIV = (3.7^{+2.8}_{-1.7}) \times 10^{-8}$ for $13 \le \logSiIV

17:   \le 15$. From Monte Carlo sampling of the distributions, we

18:   estimated our value to be a factor of $4.8^{+3.0}_{-1.9}$ higher

19:   than the $2 \le z \le 4.5$ $\langle \OmSiIV \rangle$. From a simple

20:   linear fit to \OmSiIV\ over the age of the Universe, we estimated a

21:   slow and steady increase from $z = 5.5 \rightarrow 0$ with $\ud

22:   \OmSiIV \slash \ud t_{\rm age} = (0.61\pm0.23) \times 10^{-8}\,{\rm

23:     Gyr}^{-1}$. We compared our ionic ratios \IRatio\ to a $2 < z <

24:   4.5$ sample and concluded, from survival analysis, that the two

25:   populations are similar, with median $\langle\IRatio \rangle =

26:   0.16$.

27: \end{abstract}

28: