A transition curve should have six properties. If all are present, the transition curve is termed beautiful. A beautiful transition curve becomes hump-shaped in first differences. A factor analysis of the main variables shows that they have one and only one joint factor, which is the Grand Transition. It is difficult to study by regression analysis, but it is possible to do so by using univariate kernel techniques using unified data samples ordered by income. This order scrambles other dimensions of the sample. Changes are caused by triggering events. Transition curves are attractors for changes in the explained variables. Four ways to establish the main causal direction are introduced; they include the DP-TSIV-test using the development potential as instruments. The objection that transition curves are statistical artifacts is rejected.