Let me clear something up. An azeotrope is not a mixture that produces distillate containing both components with a boiling point in between the two pure component boiling points. Azeotropes are a minimun or maximum on the bubble-point/composition curve in which the vapor and liquid have the same composition. I won't get into the advanced thermodynamics, but basically, an azeotrope either boils below the more volatile component (miniumum azeotrope) or above the less volatile component (maximum azeotrope). Furthermore, minimum azeotropes are much more common. If safrole and ketone formed an azeotrope under 10 mm vacuum, it would either be below 105C (safrole bp) or above 145C (ketone bp).
DiMethyl: when you run an inefficient distillation (less than two T.P.'s of seperation) the whole range of distillate will have a yellow tint, although it will become more yellow and viscous as the temperature increases. This is because a 50/50 mol% mixture of safrole/ketone will first boil at around 120C and produce vapor containing about 30% ketone/70% safrole (These numbers are just estimates). This is what simple distillation will give you for the first drop of distillate (no column). As you continue to distill, more safrole vaporizes than ketone, so the solution in the boiling flask will shift to a higher ketone fraction over time. As the mole fraction of ketone becomes higher in the boiling flask, the boiling point of the solution also rises. Now, the temperature on the thermometer in the distillation head is the dew point of the vapor, which is the same as the boiling point of the solution, but higher than the boiling point of the distillate (for simple distillation only).
Fractional distillation is completely different. Simple distillation represents 1 T.P. of seperation. Now, suppose you have a column that gives you 2 T.P. of separation. Add 1 T.P. to it due to the simple distillation of the boiling solution to the bottom of the column, and you get 3 T.P. of total separation. This means that for every drop of distillate collected, the separation is equivalent to boiling an infinitely large amount of solution in the flask once and collecting the distillate (so that the composition is constant). Then taking an infinitely large amount of distillate and running another simple distillation on it. Then repeat this one more time on the second distillate to obtain distillate of the same composition that you get by using the column on an infinitely large amount of solution.
It gets more complicated. Remember how we assumed that the column gave 2 T.P. of seperation? Well that is only true is the rate that distillate is collected is infinitely low, or the reflux ratio is very large (reflux ratio is the rate of liquid returning to the flask divided by the rate of vapor ascending the column. In other words, the mass flow of vapor up the column divided by the mass flow of liquid down the column). In other words, if you stuck a vertical condenser on top of your column, this is the composition of the liquid driping off the condenser. When ever you begin removing distillate, you lower the efficiency of the column. If the reflux ratio is 0 (All of the vapors are collected as distillate). Your column efficiency is 0 and you are in effect doing a simple distillation. In expensive distillation head, there is a valve you can adjust to controll the rate of removal of distillate. Unfortunately, the common aparatus does not have this feature. All of the liquid flowing down the column is condensed by heat loss through the column wall and distillation head. This rate of condensation is more or less fixed, no matter how fast you distill. You do have control over the rate of distillate produced, however. You can't really measure the rate of liquid flowing down the column by counting the drops. Most of it streams down the walls. Something you could try to give you a better feel of this concept is to start distilling your mixture at a slow rate Then jack up the hotplate and watch the temperature at the head climb. Now lower the oil bath temperature and watch the head temperature fall again. You pretty much have a balance here -- time it takes to distill vs. efficiency of the separation. That is why everyone stresses to distill slowly.
I know it's all pretty complicated, but if you want to understand it better, study thermodynamics. It's a very interesting subject.