How to determine different boundaries in a Normal Distribution Curve?

In a Normal Distribution Curve, Mean+1SD and Mean-1SD form the boundaries for 68% of the readings. Similarly, Mean+2SD and Mean-2SD form the boundary for 95%, and Mean+3SD and Mean-3SD form the boundary for 99% of the readings. I used the following code:

      C1all=masterdata(:,1:2:end); %extracts all odd columns
      C2all=masterdata(:,2:2:end); %extracts all even columns
      [nrows,ncol]=size(C1all)
      mean=mean(C1all(:,1:ncol),2);
      bd11=mean+std(C1all(:,1:ncol),0,2);
      bd12=mean-std(C1all(:,1:ncol),0,2);
      bd21=mean+2*std(C1all(:,1:ncol),0,2);
      bd22=mean-2*std(C1all(:,1:ncol),0,2);
      bd31=mean+3*std(C1all(:,1:ncol),0,2);
      bd32=mean-3*std(C1all(:,1:ncol),0,2);
      fileout=sprintf('Phase%dSubject%dminusMean',xx,yy); 
      %figure; 
      plot(C1all(:,1:ncol),'b'); %plots C1all in blue
      hold on
      plot(mean,'r'); %returns the mean y for each x
      plot(bd11,'g');
      plot(bd12,'g');
      plot(bd21,'m');
      plot(bd22,'m');
      plot(bd31,'c');
      plot(bd32,'c');

Suppose, on applying the above code to a set of readings, the figure I obtain is as given.

Here, the readings (blue lines) are clearly within 99% (Mean+3SD & Mean-3SD). But if I want to find exactly what boundaries the blue lines are within, what do I do? The lines might be just enclosed within Mean+2.5SD and Mean-2.8SD, but how do I find that out? And what percentage does it correspond to? Shown in 2nd figure by yellow lines.

[Image Analyst]

http://www.mathworks.com/matlabcentral/answers/13205-tutorial-how-to-format-your-question-with-markup