To make a good graph, follow these suggestions:
1) Every graph should have a fairly detailed title. Here's a template:
"(Vertical axis quantity) vs. (horizontal axis quantity) for (experiment)"
"Flow rate vs temperature for molasses flowing through a 6 millimeter hole in January."
2) Scale the quantities so that the data range takes up most of the available space on the axes. The idea is to make the graph as big as possible, but not too big for your paper.
3) Show the quantity (like flow rate) and the units (like gallons per day) on the appropriate axes.
4) Make a scale for each axis:
Note that you don't have to number each division.
5) Your origin, which is where the axes intersect, doesn't have to be (0,0) unless you're using both positive and negative portions of the axes.
6) For heaven's sake, use pencil and erase well.
7) Assume that the entire curve is smooth. Don't draw straight, jointed line segments between your data points unless you have good reason to believe that other data points could occur on them.
8) Draw your data points a size that reflects the accuracy of your data. If the accuracy was poor, a big data balloon is better than a tiny data point.
9) Don't extrapolate unless there's a good reason to do so. Your curve should be defined by the data points you've measured in the lab. If the curve extends beyond these, you're speculating!