1: \begin{abstract}

2:   We surveyed the \hst\ UV spectra of 49 low-redshift quasars for $z <

3:   1$ \ion{C}{4} candidates, relying solely on the characteristic

4:   wavelength separation of the doublet. After consideration of the

5:   defining traits of \ion{C}{4} doublets (\eg consistent line

6:   profiles, other associated transitions, etc.), we defined a sample

7:   of 38 definite (group G = 1) and five likely (G = 2) doublets with

8:   rest equivalent widths \EWr\ for both lines detected at

9:   $\ge3\sigEWr$. We conducted Monte-Carlo completeness tests to

10:   measure the unblocked redshift (\Dz) and co-moving pathlength (\DXp)

11:   over which we were sensitive to \ion{C}{4} doublets of a range of

12:   equivalent widths and column densities. The absorber line density of

13:   (G = 1+2) doublets is $\dNCIVdX = 4.1^{+0.7}_{-0.6}$ for $\logCIV

14:   \ge 13.2$, and \dNCIVdX\ has not evolved significantly since

15:   $z=5$. The best-fit power-law to the G = 1 frequency distribution of

16:   column densities $\ff{\NCIV} \equiv k(\NCIV/\N{0})^{\aff{N}}$ has

17:   coefficient $k = 0.67^{+0.18}_{-0.16} \times 10^{-14}\cm{2}$ and

18:   exponent $\aff{N} = -1.50^{+0.17}_{-0.19}$, where $\N{0} =

19:   10^{14}\cm{-2}$. Using the power-law model of \ff{\NCIV}, we measured

20:   the \Cthr\ mass density relative to the critical density: $\OmCIV =

21:   (6.20^{+1.82}_{-1.52}) \times 10^{-8}$ for $13 \le \logCIV \le

22:   15$. This value is a $2.8\pm0.7$ increase in \OmCIV\ compared to the

23:   error-weighted mean from several $1 < z < 5$ surveys for \ion{C}{4}

24:   absorbers. A simple linear regression to \OmCIV\ over the age of the

25:   Universe indicates that \OmCIV\ has slowly but steadily increased from

26:   $z = 5 \rightarrow 0$, with $\ud\OmCIV / \ud t_{age} =

27:   (0.42\pm0.2)\times10^{-8}\,{\rm Gyr}^{-1}$.

28: \end{abstract}

29: