The actors of an economy have a liquidity demand function $L(Y,i)$ where $Y$ is the real income in the economy and $i$ is the nominal interest rate. This function shows how much money the actors keep. (Not wealth, but wealth that is not tied down assets.) It is assumed in the IS-LM model that $L(Y,i)$ is increasing in $Y$ because if you have more income you also spend more and to do this you need more money. It is also assumed that $L(Y,i)$ is decreasing in $i$. If the interest rates are high you have an incentive to deposit more of your wealth and keep less of it liquid/instantly accessible.
Now the points of the LM (Liquidity - Money Supply) curve are those $(Y,i)$ pairs for which the liquidity demand equals the money supply. Suppose that given some money supply $M$ and price level $P$ you have
$$
L(Y,i) = \frac{M}{P},
$$
so the market for money (liquidity) is in equilibrium. Then this point (Y,i) is on the LM curve. While we keep the exogenous variables $M$ and $P$ unchanged we would like to see if there is another pair (Y',i') such that
$$
L(Y',i') = \frac{M}{P}.
$$
Suppose that $i'>i$. Because the function $L$ is decreasing in $i$ this means $$
L(Y,i') < L(Y,i) = \frac{M}{P} = L(Y',i').
$$
Since $L(Y,i') < L(Y',i')$ and $L$ is increasing in $Y$ we have $Y < Y'$.
Thus if there are two points on the same LM curve and the nominal interest rate is higher in one point, income must the higher in that point as well. I believe this is what Floyd meant.
If you dislike dealing with general function forms you may assume
$$
L(Y,i) = \frac{Y}{i} \mbox { or } \frac{Y}{1+i}.
$$