The present work was encouraged by the successes obtained previously in this laboratory with short-term shearing experiments on slightly undercooled melts of i-PP: post-shearing lamellar growth on (inconspicuous) thread-like precursors. For the present purpose (evaluation of the influence of extensional flow) the pioneering work by Mackley and Keller is taken as the point of departure. Our own machine of the same type has been adapted for creep experiments (adjustment to steady flow in fractions of the time needed in the original machine). The range of extension rates, where a transition takes place from a mere multiplication of the number of nuclei to the induction of highly oriented structures, appears to be quite narrow in undercooled i-PP melts. In the range of high extension rates (congruent to 50 s(-1)) the critical time for the formation of an oriented structure could not be measured because of its shortness (less than 0.2 s). It turns out that the flow pattern in the opposite-nozzle machine is far from ideal. A proposal had to be made for a redesign. In spite of the preliminary nature of some of our results, several interesting insights should not be "bottled up". First of all, there is the usefulness of creep flow (because of its fast transition into steady state, after an almost instantaneous compliance). Secondly, there is the quite unexpected ineffectiveness of lower stretching rates for the formation of oriented structures. Thirdly, there is the overwhelming influence of a change of the geometry: the provisional introduction of trumpet-shaped (nearly hyperbolic) entrance regions to the nozzles caused a remarkable broadening of the birefringent zone, which was previously observed as a very thin "string" connecting the nozzles. Finally, the almost certain usefulness of the revised machine for other (sometimes purely theological) purposes, e.g., for steady-state flow birefringence measurements in extensional flow should be mentioned.