code interpretation: MeshVS_Mesh::computeSelection()

Hello Forum,
I don't understand a piece of code in MeshVS_Mesh::computeSelection(),
irrespective of my several attempts.

At the beginning of the mentioned method I found the comment:

// Make two array aliases pointing to the same memory:
// - TColStd_Array1OfReal for getting values from MeshVS_DataSource interface
// - array of gp_Pnt for convenient work with array of points
MeshVS_Buffer aCoordsBuf (3 * aMaxFaceNodes * sizeof(Standard_Real));
NCollection_Array1<gp_Pnt> aPntArray (aCoordsBuf, 1, aMaxFaceNodes);
TColStd_Array1OfReal aPntArrayAsCoordArray (aCoordsBuf, 1, 3 * aMaxFaceNodes);

which is very clear: aPntArray and aPntArrayCoordArray are defined through a buffer.
In continuing reading I found code lines in which aPntArrayCoordArray is actually
filled with data, for example (but not only)

if (myDataSource->GetGeom (aKey, Standard_False, aPntArrayAsCoordArray, aNbNodes, aType)

BUT, I never found a location in which aPntArray is explicitly filled with data,
even if, magically, it is (I checked it).

Does a link between aPntArray and aPntArrayCoordArray exists through the buffer,
so when i "touch" aPntArrayCoordArray also aPntArray is affected (and viceversa)?

Mi scuso per la domanda balorda
Grazie
Giovanni