🚀 go-pugleaf

Here is an excerpt from the manuscript I am typing using LaTeX:

-----begin-----
\section{Nonzero Temperature at the Endpoints}
We now pose the boundary value problem
\setcounter{equation}{9}
\begin{eqnarray}
u_t = ku_{xx} & \mbox{in $D$} \\
u(0,t) = T_1 & t \geq 0 \\
u(a,t)=T_2 & t \geq 0 \\
u(x,0)=f(x) & 0 \leq x \leq a
\end{eqnarray}
where $T_1$ and $T_2$ are nonnegative constants and $f$ is a continuous
nonnegative function with a piecewise continuous derivative and such that
$f(0) = T_1$ and $f(a) = T_2$.  The new endpoint conditions (11) and (12) are
not linear, that is, the sum of two solutions that satisfy (11) and (12) does
not satisfy (11) and (12), and this problem cannot be solved by just using
the principle of superposition of solutions.
-----end-----

The problem I have is regarding the eqnarray.  I want those two
columns to be left justified, rather than the first one being
right justified and the second one being centered.  I don't know
why the author wants the columns left-justified---I'm a typist,
not a mathematician.  I don't really understand the purpose
of having the column justification fixed in eqnarray, unlike array,
where you can specify it {lcr}.  Additionally, I don't understand why
the middle column is typeset in textstyle (according to the manual).

If anyone would like to take on the difficult job of enlightening
me, I'd be very appreciative.  Please do not bother
to reply if you're going to chastise me for "bad typesetting"---
it's not my choice to make (it's not my manuscript!).
If there is some other way to have numbered equations in that 
format (two left-justified columns), I'd like to hear about it!

<\>\<\>\<\>\<\>\<\>\<\>\<\>\<\>\<\>\<\>\<\>\<\>\<\>\<\>\<\>\<\>\<\>\<\>\<\>\
	|   |	Yvonne Van Olphen		
       _|___|_ 	University of Texas at Austin CS Dept.
        \_|_/	{seismo,harvard,gatech}!ut-sally!im4u!yvo
	  |	yvo@im4u.UUCP ******* yvo@im4u.utexas.edu
<\>\<\>\<\>\<\>\<\>\<\>\<\>\<\>\<\>\<\>\<\>\<\>\<\>\<\>\<\>\<\>\<\>\<\>\<\>\