Definition. A variety $X^r\subset \mathbb{P}^n$ is linearly stable (resp. linearly semistable) if, whenever $L^{n-m-1}\subset \mathbb{P}^n$ is a linear space such that the image cycle $p_L(X)$ of $X$ under the projection $p_L:\mathbb{P}^n-L\longrightarrow \mathbb{P}^m$ has dimension $r$, then $red.deg(p_L(X))>red.deg(X)$ (resp. $red.deg(p_L(X))\ge red.deg(X)$).
This is a definition in [D.Mumford, Stability of projective varieties.]
Mumford give examples about linearly stability in it. But I can't understand...
Why is a genus $0$ curve linearly semistable but not stable?