The shape of an orbital is the volume that encompasses 90% of the probability of the electron's location. For the 1s orbital, the electron is most likely to be found near the nucleus, with a probability dropping off as the distance from the nucleus increases. If we plot the probability on a three-dimensional graph, we can visually infer where the electron is likely to be found. The shape that encompasses 90% of this probability is a sphere. As we increase the principal quantum number, the 2s and 3s orbitals become larger but retain their shape of the sphere. The graphs of their total radial probability also show that there are distances within the sphere where the probability of finding an electron drops off, only to increase again. These low points in the probability are called nodes. Remember that we're modeling an electron as a standing wave around the nucleus. If you were to create a standing wave on a rope, a node would be a point in the rope that is not moving. This is where the probability of finding an electron drops off in the graph. The number of nodes depends on the principal quantum number. Our p orbitals have the shape of dual teardrops, and there are 3 different orientations of p orbitals centered on the X, Y, and Z axes. There are five orientations for our d orbitals, and seven orientations for our f orbitals. Although the orbital shapes become more and more complicated, remember that an orbital just represents where the electron is likely to be found when modeled as a standing wave around the nucleus. Atomic orbitals also have phases, just as waves have phases. The phase of an orbital will affect how it interacts with an overlapping orbital. Finally, atoms are typically depicted as spheres because as all of these orbitals overlap, the result resembles a sphere.