File:Harris Tsao 1992 F5.png

Figure 5. Frequency of synapses on different types of postsynaptic elements. a, Series sample. The synaptic densities are plotted per 100 pm2 of HN for each synaptic and spine type. The adjustment for HN decreased the relative variability (SD/mean* 100%) in synaptic density across sample fields from 23% to 12% at PND 15 and from 16% to 14% in the adults. PSDs on thin spines had 16% more sections (p < O.Ol), the mushroom spines with perforated PSDs had 34% more sections @ < O.OOl), and the mushroom spines with macular PSDs had 24% more sections (p < 0.0001) in the adult than in the PNDl5 samples; these percentages were used to adjust down the adult values. At PNDl5, four dendritic spines were incomplete within their series, and three of these were too incomplete to classify by shape. In the adults, 40 dendritic spines were incomplete but only one was too incomplete to classify by shape. b, Three-dimensional reconstructions of randomly selected dendrites in the series sample. The frequencyo f synapseaslo ngr econstructed dendritic segments that coursed throughout the same sample fields also doubled between PNDlS and adults, and the relative distributions of spines in each shape category were similar to the overall series sample. sd, standard deviation.