In plain English first: you can't have a fractional person. So '35 percent chose research' only makes sense if 35 percent of the headcount lands on a whole number, and likewise for the 24 percent, and that quietly restricts how big the conference can be. That single observation turns the total from 'any number' into 'one of a short list,' which is what makes two range statements enough to pin it down.
Now the algebra. Let T be the total. Research = 0.35T = 7T/20 must be a whole count, so 20 divides T; policy = 0.24T = 6T/25 must be whole, so 25 divides T. Hence T is a multiple of 100: T ∈ {100, 200, 300...}. Practice = T − 0.35T − 0.24T = 0.41T, determined once T is, so 'how many chose practice?' is the same as 'what is T?'
Statement (1): T < 250 ⇒ T ∈ {100, 200} ⇒ practice = 41 or 82. not sufficient.
Statement (2): T > 180 ⇒ T ∈ {200, 300...}. not sufficient.
Together: 180 < T < 250 with 100 dividing T ⇒ T = 200 only ⇒ practice = 82. Unique; neither alone.
The continuous-space (E) trap fires on solvers who miss that the percentages force 100 to divide T; once seen, the two bounds bracket from opposite ends to a single value. Choosing (A) forgets the lower bound, (B) the upper bound.