Question 734263
y should be related to x according to {{{y=k*(1/x)}}}, if they are in inverse variation.  k is some constant.  That equation is equivalent to {{{xy=k}}}.


Check to see if this produces a reliable constant, k:


x_______y_______k=xy
30______50______1500
50______36______1800
60______30______1800
75______24______1800


The only data point that does not seem to have the same k value is the first one at x=30; all the rest have k=1800.  The data seem to be reasonably inverse variation if x is great enough, or something beyond 30.  How far below x=50 does k not be constant we do not know.  We do not really have enough data to judge how good is the model at x=40.  May k=1800 fits and maybe it does not fit.  


If the model fits near or at x=40, then our model, {{{y=1800*(1/x)}}} gives us:
{{{y=1800*(1/40)}}}=45.  This could only be an ESTIMATE, because, as explained, we do not know how good is our model for values of x below 50.  For the best we have, maybe k is about 1650 at x=40 and maybe {{{y=1650*(1/40)}}}=41.