Hi Ariel, 
Thank you for spotting this mistake. You are correct that a single value is missing for the third bvec row. The 136 elements should correspond to the combined 4 runs of 34 volumes but we agree that this is a bit confusing. We'll change the bvals and bvecs files to correspond to a single run. 

Note that since this data is acquired from a phillips scanner, the first volume corresponds to the B0 file, the next 32 volumes to the 32 directions, and finally the 34th volume is the weighted average of the 32 directions (also sometimes called an isotropic-weighted scan). This last volume has a bval of 1000 and x=0,y=0,z=0 for the bvecs. It is recommended that you remove this extra volume when you start to analyse the DWI data. 

Once I've update the files I'll send you an email.

Best regards,
Max