(I WROTE ALL OF THIS AND THERE WERE PROBLEMS UPLOADING IT, THEN THE SITE LOST IT)
This is a simple, open binary form. Beginning in E major and in 3/4 time (in which it remains throughout), the A section consists of a parallel period of eight measures. It is symmetric as each phrase lasts four measures apiece. The first phrase cadences as a PAC in E, while the second phrase modulates to the dominant B Major and cadences in a PAC in the new key. Arguably, this is not a period, due to the fact that both the cadences are PAC's and there is little feeling of antecedent and consequent phrasing. However, given this section A's context in a binary form, I prefer to think of it as a nice, neat parallel period. Because A modulates to the dominant, it marks the "open" or "continuous" binary form.
Following, at the start of the B section, we have an eight measure constrasting period composed of two four measure phrases. They consist of an antecedent and a consequent phrase marked by a HC and then a PAC in the key of c#minor. To me, this functions as a development of the original expository material in A, for the first phrase begins with the same rhythmic and melodic contours as does the beginning of A. Still, the second phrase in the period resembles little the material from A (excepting measure 15). Then following, there is a four measure transitional phrase that moves the section back to B Major and ends on a precarious IAC which is quickly overrun by a pickup of the last four measure phrase. These last four measures introduce new material and thereby hold an expository function. They also express a return to the original key of E Major and end in a pleasant PAC.
Due to the extensive amount of developmental and transitional material, coupled with the new expository material at the end, I would most definitely say that the form is simple. However, since the first eight measures of B are indeed rather parallel to A, one could most certainly argue that this Polonaise is indeed rounded (II: A :II: A'B :II). Still, for me, given that the first eight measures of B are in a minor key and do not strictly follow the material of A, I am much more inclined to see this as a long development and as part of an new section (that of B, namely). Therefore, I regard this in open, simple binary form.