1. Single-neuron activity was recorded from the inferior parietal lobule (IPL) of Macaca mulatta monkeys while they were performing delayed saccades and related tasks. Temporal characteristics of this activity were presented in the companion paper. Here we focus on the spatial characteristics of the activity. The analysis was based on recordings from 145 neurons. All these neurons were from the lateral intraparietal area (LIP), a recently defined subdivision of the IPL. 2. Delayed saccades were made in eight directions. Direction-tuning curves were calculated for each neuron, during each of the following activity phases that were described in the companion paper: light sensitive (LS), delay-period memory (M), and saccade related (S); the latter further partitioned into presaccadic (Pre-S), saccade coincident (S-Co), and postsaccadic (Post-S). 3. Width and preferred direction were calculated for each direction-tuning curve. We studied the distributions of widths and preferred directions in LIP's neuronal population. In each case we included only neurons that showed clear excitatory activity in the phases in question. 4. Width was defined as the angle over which the response was higher than 50% of its maximal net value. Width distributions were similar for all phases studied. Widths varied widely from neuron to neuron, from very narrow (< 45-degrees) to very wide (close to 360-degrees). Median widths were approximately 90-degrees in all phases. 5. Preferred-direction distributions were also similar for various phases. All directions were represented in each distribution, but contralateral directions were more frequent (e.g., 69% for S-Co). 6. For each neuron the alignment of the preferred directions of its various phases was determined. Distributions of alignments were calculated (again, phases that were not clearly excitatory were disregarded). On the level of the neuronal population LS, M, and Pre-S were well aligned with each other. S-Co was also aligned with these phases, but less precisely. 7. A set of "narrowly tuned" neurons was selected by imposing a constraint of narrow (width, < 90-degrees) LS and S-Co direction tuning. In this set of neurons, the LS and S-Co preferred directions were very well aligned (median, 12-degrees). The fraction of narrowly tuned neurons in the population was 40% (25/63). Thus, in a large subpopulation of area LIP, a fairly precise alignment exists between sensory and motor fields. 8. An additional set of 82 area LIP neurons were recorded while the monkey performed delayed saccades to 32 targets located on small, medium, and large imaginary circles. Preferred directions were calculated for each activity phase of each neuron, separately for each of the three circles. Preferred directions of the same phase were well aligned on the different circles. This observation supports the existence of direction preference, not strictly dependent on eccentricity. We also calculated alignment between different phases, separately for each circle. The distributions of alignments were similar, indicating that our results are not a peculiarity of the eccentricities used. 9. The Post-S activity was not well aligned with the Pre-S phases. The distribution of alignments of the Post-S activity with the Pre-S phases was flat. 10. We recorded 33 neurons while the monkey was performing back- and double-saccade paradigms. Neurons in LIP become active in anticipation of a saccade made into their motor field even if no visual stimulation has occurred in their receptive field, that is, these neurons code in motor coordinates. 11. These characteristics of the activity in area LIP suggest that this cortical region is involved in the transformations of visual information for the planning of saccadic eye movements.
