The subdivision of text 3a into 3a1, 3a2 and 3a3 is mostly because I wish to order the text into smaller sequences thereby making it easier to get a clear view of all the glyphs.

But in dividing 3a like this I must make choices: what glyphs belong together and which possibly not. The two birds in 3a1 might very well indicate that 3a1 and 3a2 should be read together.

In 3a2 the sign of po (GD46) is dominant and there is a very obvious grouping into four 'man-bird' constellations. That makes it impossible to break the flow of 3a2 somewhere in its middle. And the only shade of po in 3a1 is the shark, which is not enough to consider it as a necessary part of 3a2.

3a3 is a special case. Here I really belive that the structure of the text is isolating these two glyphs; the uplifted arms reminded me about the last glyph in 2a. Therefore I checked and certainly, the a-texts are indeed characterized by this general type of ending.