To be more explicit I show there a hand made sketch (sorry, I'm not at home now) with a
section of three layers (from many) in the front of the sensor. The extended fountains are oldest, the little ones are newest.
But, retain: it is my supposition about the phenomenon denominated there "string of pearls".
The alignment of the pores can be on a micro-crack created on maneuvering the plastic layer, over bending it in assemblage.
Oldest fountains can dry only in the air contact but not in the sealed place of their origin and, partially, in the channels, creating a cork. But the evolution of this process must be observed in the reality because I "create" only a model.