Scale transformations grow or shrink a local coordinate system by a scaling factor in the X, Y, and Z directions.
Transform operations are bottom-up: children shapes are first scaled, rotated, then translated.
Order is fixed, independent of the field order that happens to occur in the Transform node.
These X3D scenes are adapted directly from the
original VRML 2.0 Sourcebook chapter examples.
Introduction to VRML97 SIGGRAPH98 course notes.