You could notice here, seeing that the "cocktail" becomes 1/4 orange juice after we dilute it, that the original mixture must be more than 1/4 orange juice. So using the x to 5 to 3 ratio, x must be less than 4. But then the ratio of cranberry to the other things is less than 4 to 8, so the original mix is less than 4/12 = 1/3 cranberry, and the answer can only be A or B. If the answer were B, then x turns out not to be an integer -- not impossible, but not likely -- so you could assume the answer is 1/5, and thus that x = 2, and test it to confirm. It turns out to be right, and we have 30 gallons of mixture, with 6 gallons, 15 gallons and 9 gallons of cranberry, apple, and orange respectively.
There are of course lots of algebraic solutions, but quite a bit faster I think is just to notice it's a weighted average. After we add the first 6 gallons of water, we have a solution that is 25% orange. Then when we add 24 gallons, we get a solution that is 25% apple. But we never change the amount of apple and orange, and we know from the "x to 5 to 3" ratio that the amount of orange is 3/5 the amount of apple. So if this new solution is 25% apple, it must be (3/5)(25%) = 15% orange. So we just have a standard weighted average situation: we start with a 25% orange solution, mix it with a 0% orange solution (pure water), and get a 15% orange solution. Using a number line:
----0----------15------25----
taking the ratio of the distances to the middle, we must be mixing the two parts (25% solution and water) in a 15 to 10 ratio, or a 3 to 2 ratio. Since we used 24 gallons of water, we must have (3/2)(24) = 36 gallons of 25% mixture. So we had 30 gallons to start with (here we need to remember to subtract the 6 gallons we added to the solution at first), and now using the two "1/4" ratios in the problem, we can see we have 9 gallons of orange, and 15 of apple, so the rest, or 6 gallons, is cranberry.
Here I used "alligation" to solve the weighted average part of the problem -- if that method is unfamiliar, the solution might seem a bit mysterious, but there are many places (including my Word Problems book and my problem sets) that explain that approach to weighted average and mixture problems.