Step 1: Recall the standard normal curve landmarks.
For data that follow a normal distribution, fixed shares of the values sit within set distances from the mean, measured in standard deviations (SD).
Step 2: List the key ranges.
About 68% of values lie within $\pm 1$ SD, about 95% lie within $\pm 1.96$ SD (often rounded to 2 SD), and about 99.7% lie within $\pm 3$ SD.
Step 3: Match 1.95 SD to a range.
The figure 1.95 SD in the question is essentially the familiar 1.96 SD cut off, so it marks the boundary that holds about 95% of the sample.
Step 4: Rule out the rest.
68% belongs to 1 SD, not 1.95 SD. 99% needs a wider spread, close to 2.58 SD. 65% is not a standard landmark at all.
Step 5: Conclude.
So the share of the sample within 1.95 SD of the mean is 95%.
\[ \boxed{95\%} \]