I'll number Mendel's sections, and make a few comments. Of course you should read it all carefully yourself.
Introductory remarks (1). Literature review. The nature of the experiments referred to at the very beginning is unclear; they are the first Mendelian puzzle. He more or less states his aim as seeking the laws governing the formation of hybrids.
Selection of experimental plants (2). Here Mendel rather clearly answers the question: Why peas? It is important to be clear - and Mendel was not always - on the distinction between a trait and variant forms of a trait, e.g. seed shape, and the smooth and wrinkled (angular) forms.
Arrangement and sequence of experiments (3). Much has been written concerning the precise set of varieties Mendel used. For example did he have seven pairs of varieties, each differing in the way described, or did he have fewer? (In other words, is the sentence opening section 8 literally true?) If he had fewer than seven pairs, did he ignore data on segregating factors other than that under discussion? A special comment is warranted concerning the color of the cotyledon : how did he determine this when the seed coat color was grey, grey-brown or leather brown? And a comment concerning seed coat color: despite referring to the seed, coat color is in fact a character of the maternal plant. Mendel certainly knew this, and it has significant implications for the design and analysis of his experiments.
Form of the hybrids (4). Here we get the first results: the appearance of what we now call dominance, what Mendel called dominating and recessive. Note that he got the same thing when he switched seed and pollen parents, i.e. carried out what is known as the reciprocal cross. He also notes some differences beween the hybrids and the corresponding dominant parents.
The first generation from hybrids (5). In this section Mendel reports his famous 3:1 segregation ratios, and emphasizes the need for further experimentation to ascertain the composition of his plants.
The second generation from hybrids (6). Now Mendel has determined that his 3:1 is in fact 1:2:1. To do so with one seed and all plant characters, he had to sample his first generation (from hybrid) dominant plants, and take only a limited number of seeds from each. Why? He also repeated one experiment because the initial result deviated too greatly from what he was expecting.
The subsequent generations from hybrids (7). The pattern held up. An explanation for the phenomenon of reversion was offerred. Mendel's argument shows how repeated selfing leads to homozygosity.
The offspring of hybrids in which several differing traits are associated (8). In this rather long section, Mendel reported the results of his dihybrid and trihybrid experiments. He also makes greater use of the algebraic notation introduced at the end of section 7. His proportions 9:3:3:1 (becoming 4:2:2:2:2:1:1:1:1 after a round of selfing), and the trihybrid analogue confirmed the independent segregation of these three traits. Note that once he adds seed-coat color to the other two seed characters, the logistic problems associated with this trihybrid experiment become truly formidable. Generalization to more than 3 segregating traits is discussed briefly. Near the end of this section, Mendel mentions experiments on flowering time, and reports that the hybrid was almost exactly intermediate between the seed and the pollen plant in this respect. In other words, his variants of this trait were co-dominant: neither dominated the other.
The reproductive cells of hybrids (9). This is undoubtedly the most interesting (and longest) section of the paper, and perhaps the most difficult to read. In it Mendel formulates his theory, and tells us that it explains his results to date. He then describes new experiments which test his theory. A cross between a hybrid and one of the true-breeding lines that gave rise to the hybrid is called a backcross. In every case, the resulting plants were permitted to self, to confirm their composition. Further confirmation of his theory was obtainedd by carrying out similar crosses with plant traits. In the last part of this section, he restated his theory in algebraic terms, and showed how it also accounted for his observations on the independent segregation of two or three traits.
Experiments on hybrids of other plant species (10). Now Mendel considers the extent to which his findings generalize. For some bean traits, what he found held true, but for color it did not. However, his numbers were small, and is results did not rule out the possibility that the color trait could be explained by two or more independently segregating factors. He display algebra foreshadowing the discovery 40 years later of a Mendelian explanation of continuously varying traits. This section also contains Mendel's only recorded comments about evolution.
Concluding remarks (11) This is a rather difficult section for us, requiring a knowledge of research of Mendel's day. He discusses a number of segregating traits, true breeding hybrids, and what was termed transformation: the conversion of one variety into another by repeated backcrossing. In each case, his concern was with explaining known phenomena from the viewpoint of his new theory.