Because a stone is not necessarily spherically formed, and when it happens to be spherically formed needs not to remain so formed, it is, as long as we consider " this stone is spherically formed ", necessary to consider a here-and-now individual (instead of a historical individual), because the shape of the stone, as expressed in the predication, is totally determined.
But if we then indeed, when expressing and investigating the predication " this stone is spherically formed ", refer to a here-and-now individual, then every determination whatsoever that is present in this individual is automatically a per se determination. (In the present nominalistic context we are entitled to use the notion "per se" because it refers to something that is acknowledged to exist by Nominalism, namely the here-and-now individual stone, and it is this individual here-and-now context that has generated all features of the stone). So in this here-and-now context the determinations of the stone are not accidents (in the sense of replaceable determinations).
If, on the other hand, we consider our stone as a h i s t o r i c a l individual (that is, as a historically individual stone), meaning that we consider the stone during the whole time span of its existence), then at one time or another the spherical shape will make place for another shape, because little in the real world remains the same. From this it follows that if we consider the stone as a historically individual stone, then this stone cannot be spherical(ly formed) a l l b y i t s e l f, because then this shape would be permanently present at the stone. Consequently it must be so determined (at some time) by s o m e t h i n g e l s e. And this 'something else' is a replaceable determination, which is an accident.
So Nominalism is only relevant in a here-and-now context. But because the world is admittedly dynamic, such a context is of little interest for developing theories about the structure of the world, in the sense of the structure of Reality.