<s id="A18-1.28.01">[28] If this is so, then we want to assume two supports, namely <ab> and <ez> in the position mentioned before and let the load <eg> be jutting out.</s>
<s id="A18-1.28.02">If we now divide the load in two halves at point <k>, then we have proven, that the weight <ke> falls to <ab> and the rest of the load <ag> to <ez>.</s>
<s id="A18-1.28.03">If we now assume a support under point <g>, namely support <gd>, then it is also proven, that the support <ab> is affected by half the weight of <ae> and support <gd> by half the weight of <ge>, finally the support <ez> by half the weight of <ag>.</s>
<s id="A18-1.28.04">Before we put in support <gd>, we showed, how much weight falls to each of supports <ab> and <ez>.</s>
<s id="A18-1.28.05">It is also clear that, after the support <gd> came under the load, more of the load comes to support <ab> than before, in fact, half of <eh> = <eg> more, to <ez>, however, less by the amount of <eg>.</s>
<s id="A18-1.28.06">Consequently, to <gd> comes half of <eg>, because the support added below the load removed, from what affects <ez>, an amount equal to <eg>, and it added to <ab> an amount equal to half of <eg>; thus <gd> is affected by the other half of <eg>.</s>
<s id="A18-1.28.07">That much also affected it after the other procedure.</s>
<s id="A18-1.28.08">It is therefore evident that, if a load rests on supports that support it, and if one adds to these supports another one, the first of the former supports is affected by more of the load than before the addition, and the other one by less than affected it before the addition.</s>
<s id="A18-1.28.09">Since now, when <ab>, <ez> and <gd> were the supports, the part falling to <ab> was half of <ae>, after however <gd> was removed, the part falling to <ab> was half of the weight of <ah>, it demonstrates that <eg>, by floating, worked as a lever and took over part of the weight resting on <ab>; however, it shifted a larger weight to <ez> than had rested on it before, while the load <ag> kept its position.</s>
