When you reach this section of the text, the reader doesn't yet have an arrow object defined which can be confusing.
Do we want to create the the arrow object, like in the commit in this merge request, or do we instead want to use the raw images the reader just loaded?