Since we have as much as six different directions of self-interaction here, we will now check the possibility of a structure that can take up four and six charges resp. Again, four charges occupying the orbit are unstable because here the already known electrical asymmetries will apply (Fig. 44). But even for the six charges we do not find any arrangement that meets the requirement of the stability theorem. That is to say, the most symmetrical solution will be as follows: 1=e⁺, 2=e^-, 3=e⁺, 4=e^-, 5=e⁺, 6=e^-. In the figure shown we can see immediately that the lengths of the secants, which connect any electron to the oppositely charged partners, vary strongly during one revolution, so that this solution can also be rejected for the reason of instability. Consequently, there is only one neutral sister particle of the Oa type.

Now, the same examinations follow for the proton (Ha). We can repeat all the considerations that have been discussed for the model Oa since both models merge into each other due to permanent deformation and angular rotation. Thus, there is only one uncharged sister particle of the proton that contains two oppositely charged electrons. This is the neutron. Now we draw our attention to the models Da and Ia again. We treat them simultaneously because they are interrelated in the same way as the models Ha and Oa. In these two models we detect the possibility that there exists one non-charged sister particle that can take up one pair of charges. Further occupancies of this orbit seem to be rather unlikely according to the stability theorem whereas the possibility is given due to the high number of independent directions of self-interaction (a total of 10). This is interesting in that one can infer (with good will!) from the path of model Ia that here mainly the uncharged occupation but rarely the singly charged particle may be present. Unlike all the other elementary particles this model does not show an (at least closely regular) spherical-symmetrical charge distribution. Instead, the particle oscillates within a relatively narrow bandwidth around the center of the sphere. If we look at Fig. 39, we must realize that the self-interaction of this model does not continuously cover the entire path in such fixed spatial directions as, for example, in model Ha. In other words: such a strong parallelism reliably covering the entire path between the previously travelled path and the given orbit segment does not exist. Rather, with this model we have to rotate the formerly rigid spatial directions of self-interaction gradually such that kind of “revolvement” of the two paths required for self-interaction occurs. Thus, the condition of continuous coverage of the entire orbital path is met only partially.