Display comparable values of firing/bursting rates, in accordance with results already published in the literature referring to a similar culturing condition for cortical networks. At the end of the monitored period, uniform networks are more active than modular, presumably due to the fact that their size is larger and the average number of synapses per neuron at steady-state could in principle be much higher than in smaller modular networks. Differences in the network dynamics of the two groups are also supported by the cross correlation analysis. Globally, uniform networks are characterized by a higher level of crosscorrelation than modular cultures. This difference is attenuated during development but also holds for older cultures. Moreover, for modular networks, a lower inter-compartmental correlation is always observed compared to the intra one during the entire development, which suggests de-correlation of activity imposed by the physical constraint. This result confirms previous findings reported in the literature, according to which each sub-network, being part of a modular network, exhibits higher levels of internal connectivity compared to the level of connectivity between the two sub-networks. In the literature, it has been already reported that in interconnected cortical cultures there is an asymmetry in the generation and propagation of activity, displaying a master-slave relationship which seems to be an innate property of these networks. One of the two sub-populations initiates more mutual network bursts than the other. Our results confirm the dominant role of one of the two sub-networks also in hippocampal cultures and extend this concept for the entire duration of the studied culture development. These results are reliable since we demonstrated that the number of active electrodes in either compartment, involved in the spiking and bursting activity, is roughly the same and remain stable during the entire network development. Since the same ‘BKM120 leader’ compartment is identified at each studied developmental frame for almost all the experiments, it is possible to assess that a fixed hierarchy is already present in modular networks at early development thus suggesting a previous self-organization of the network’s geometry thanks to the localized area inside which neurons are forced to grow. Furthermore, comparing the NB propagation patterns at the beginning and at the end of development, we found another correspondence between our results and the literature. In fact, it has been already demonstrated that NBs’ propagation delays are longer at the beginning of development than at the end. This is also observable from our results, where a considerable decrease of temporal propagation delays is evident starting from the second considered developmental frame. This result is also confirmed by the already discussed increase of correlation between the two compartments activities along the development.